https://publicacoes.softaliza.com.br/cilamce2025/issue/feed 2025-12-18T20:47:21+00:00 Open Journal Systems https://publicacoes.softaliza.com.br/cilamce2025/article/view/13619 2025-12-18T20:45:00+00:00 GLEIDE LINS gleidekarolayne@lccv.ufal.br Catarina Nogueira de Araújo Fernandes catarina@lccv.ufal.br Thiago Barbosa da Silva thiago.barbosa@lccv.ufal.br William Wagner Matos Lira william@lccv.ufal.br Charlton Okama de Souza charlton@petrobras.com.br

This study proposes a thermo-poroelastic numerical modeling approach to evaluate the structural integrity of the cement sheath in oil wells. Positioned between the casing and the surrounding formation, the cement sheath is responsible for casing support and hydraulic isolation between permeable zones throughout the well’s lifecycle. Although thermoelastic models are commonly used in design practice, recent studies highlight that the porous behavior of hardened cement significantly affects its mechanical response under thermal and pore pressure variations. Analytical models provide valuable preliminary insights, but rely on simplifications that limit their applicability to real-world conditions. In contrast, numerical modeling enables more accurate representations of operational scenarios and stress distribution in the casing–cement–formation system. In this context, the methodology includes: (i) idealization of the physical model; (ii) characterization of the case study based on literature data; (iii) thermoelastic and thermo-poroelastic numerical modeling; and (iv) comparison of numerical results with analytical solutions. The results show excellent agreement with analytical predictions, with relative errors below 1.5\%, confirming the accuracy of the adopted numerical approach. Therefore, the proposed numerical model represents a significant advancement in the structural integrity assessment of cement sheaths, supporting safer and more efficient well designs.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13635 2025-12-18T20:45:43+00:00 Renata Corrêa redrummondc@outlook.com Bruna Lima Veras Maia brunalimaveras@gmail.com Eduardo Pescada eduardopescada@gmail.com Bruno Furieri bruno.furieri@gmail.com Taciana Toledo de Almeida Albuquerque tacianatoledo26@gmail.com

Open pit mining activities associated with the action of wind on exposed áreas can significantly influence local air dynamics, contributing to the emission and dispersion of particulate matter that affects air quality and human health, especially particles smaller than 2.5 micrometers, which are capable of penetrating the entire respiratory system. Computational Fluid Dynamics (CFD) is a powerful tool that allows for the analysis of airflow patterns over an open pit mine and the prediction of areas with higher potential for particulate matter emission. Unlike previous studies that use simplified terrain geometries, this research integrates high-resolution topographical data obtained from official surveys, ensuring greater accuracy in representing real surface features. Similar studies have shown that realistic topography plays a role in improving the stability of the results with respect to mesh refinement, particularly in regions os steep terrain.The simulations were conducted using Ansys Fluent, employing the Reynolds-Averaged Navier-Stokes (RANS) turbulence model with the k-epsilon closure to capture turbulent flow characteristics. The computational domain was discretized with a high-resolution mesh of approximately 12 million elements, incorporating wind velocity profiles derived from a nearby weather station to reflect realistic atmospheric conditions. Results highlight zones of high and low wind velocity over the pit and surrounding areas, indicating potential regions for particulate matter accumulation or dispersion.Overall, this work demonstrates the effectiveness of combining high-fidelity terrain data with advanced CFD modeling to improve the understanding of airflow in mining environments. The methodology supports improved planning and implementation of mitigation technologies aimed at reducing atmospheric pollution caused by mining activities. Consequently, this research contributes to enhancing air quality management and protecting public health in mining-affected regions.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13651 2025-12-18T20:46:19+00:00 Juan Carlos Cabral jccabral@pol.una.py Gustavo Espínola gustavoespinola@fpuna.edu.py Jhabriel Varela jhabriel@pol.una.py Christian Schaerer cschaer@pol.una.py

Accurate and efficient resolution of sparse linear systems arising from computational fluid dynamics simulations is a computationally intensive task. Traditional iterative methods such as GMRES(m) often suffer from stagnation when applied to large, non-symmetric matrices derived from discretized fluid dynamics models. To address this limitation, a modified version of GMRES(m), known as LGMRES(m,l), was developed to incorporate additional error vectors in the search space. However, previous studies have mainly focused on increasing the subspace dimension m while keeping the number of error vectors l fixed. From a computational cost perspective, this approach may be inefficient, as larger subspaces demand increased memory and processing time, without necessarily enhancing convergence rates.To address this gap, the present work proposes an adaptive enhancement to the LGMRES(m,l) method, incorporating a control-based strategy for dynamic adjustment of both parameters. A proportional control mechanism adjusts m based on the observed rate of convergence, allowing increased subspace dimensions in phases of slow convergence and reduction during fast convergence. Currently, the parameter l is dynamically updated by evaluating the magnitude of error vectors, retaining only those that contribute significantly to residual reduction. This dual adaptation framework aims to mitigate stagnation while optimizing computational resources.The proposed adaptive LGMRES is tested on matrices derived from computational fluid dynamics applications, including groundwater flow models, contaminant transport simulations, and atmospheric modeling for validation. Benchmark matrices from the SuiteSparse Matrix Collection specific to these applications will be used to assess the robustness and efficiency of the adaptive LGMRES method compared to its static counterpart and other iterative solvers. Numerical results obtained using benchmark matrices derived from environmental fluid dynamics applications confirm the effectiveness of the proposed adaptive LGMRES. Significant reductions in the number of iterations and the overall execution time were observed, highlighting the potential of adaptive parameter tuning to improve the robustness and scalability of LGMRES in contexts of computational fluid dynamics.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13667 2025-12-18T20:46:49+00:00 Gino Bertollucci Colherinhas gino@ufg.br Hisham Tariq 100063744@ku.ac.ae Marcus Vinicius Girão de Morais mvmorais@unb.br Francesco Petrini francesco.petrini@uniroma1.it Agathoklis Giaralis agathoklis.giaralis@ku.ac.ae

This study presents the optimal structural redesign of the IEA-15MW offshore wind turbine tower using a single-objective genetic algorithm coupled with a high-fidelity finite element model implemented in PyMAPDL. The optimization aims to minimize the tower mass while satisfying stress, frequency, displacement, geometric, and buckling constraints under site-specific environmental conditions. Four representative Design Load Cases (DLCs) were considered—normal, extreme, and idling sea states—based on wind and wave spectra typical of Northern Europe. Spectral dynamic responses were computed via power spectral density (PSD) analysis and applied in combination with static wind and wave loading. All optimizations converged to a single robust structural configuration, leading to a 9.94\% mass reduction compared to the IEA-15MW reference design and a reduction in first natural frequency from 0.173 Hz to 0.140 Hz, satisfying the dynamic regime. Post-optimization buckling verification was performed based on Eurocode 3 Part 1--6, confirming shell stability under combined axial and shear stresses. The proposed framework demonstrates the potential of customized GA-based workflows for efficient and constraint-aware offshore tower design.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13626 2025-12-18T20:45:18+00:00 José Ribeiro telles@eng.uerj.br Americo Cunha Jr americo@lncc.br

Polymer-based plastics such as acetal exhibit time-dependent deformation under constant stress, known as creep, which can eventually lead to rupture or static fatigue. A common misconception is that materials under tolerable static loads remain unaffected over time. However, accurate long-term deformation predictions require experimental data and advanced modeling techniques. Traditional rheological models, composed of basic elements like springs and dampers, often fall short in capturing the intrinsic power-law behavior of creep. While the springpot—based on fractional calculus—provides a power-law relationship, its fixed-order nature limits its effectiveness when the deformation rate evolves over time. This study presents a nonlinear creep model based on a variable-order springpot, specifically developed to characterize the viscoelastic behavior of acetal. The model dynamically adapts to the evolving properties of the material, accurately capturing transitions among the glassy, transition, and rubbery phases. Model parameters are calibrated using a robust identification procedure based on the cross-entropy method, resulting in physically consistent and highly accurate predictions. This advanced modeling framework not only overcomes the limitations of fixed-order formulations but also establishes a foundation for the application of variable-order mechanics to viscoelastic materials—offering a powerful and tailored tool for predicting the long-term structural performance of acetal.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13642 2025-12-18T20:45:56+00:00 Luiz Antônio Malheiros Filho luiz.malheiros@ufba.br Castro Baptista Elias castroelias@ufba.br Alexandre de Macêdo Wahrhaftig alixa@ufba.br

Column stiffness is a fundamental parameter for the dynamic analysis of buildings. One of the mostimportant applications of structural dynamics theory is the use of the equivalent system concept, which allowstransforming systems with infinite degrees of freedom into one containing only a single degree of freedom. Withthis approach, multi-story buildings can be modeled as single-degree-of-freedom systems, allowing the stiffnessand mass of the columns to be included in the analysis. In the approach developed in this work, a standard columnis assumed to compose the floors of a building, with its properties being calculated based on the generalizedcoordinate assumed for its boundary conditions. In this context, the column's stiffness is calculated to take intoaccount its geometry, the influence of concentrated and distributed axial loading, and a certain degree of rotationalrestriction at the base, considered by the application of a rotational spring located at this position. The simulationassumes an arbitrary number of columns per floor and a certain number of floors in the building. A threedimensionalsystem is therefore represented by an equivalent system obtained by combining springs and masses,first in parallel, representing the grouping of columns on the floors, and then in series, representing the associationof the floors within the building. With this strategy, it was possible to determine the equivalent stiffness of eachfloor as the sum of the column stiffnesses, and that of the floors as the inverse of the sum of the inverses of thefloor stiffnesses at the different levels. The equivalent mass is obtained by eigenvalues in relation to the seriessystem. Thus, the natural frequency of the original model could be calculated using the concept of an equivalentsystem calculated under the influence of the loading, making this a non-linear approach.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13658 2025-12-18T20:46:32+00:00 AMANDA DE SOUSA MACHADO amandasousarj@gmail.com Fernanda Lins Gonçalves Pereira fernanda.pereira@eng.uerj.br Rodrigo Bird Burgos rburgos@eng.uerj.br José Guilherme Santos da Silva jgss@uerj.br

The current trends in structural design pointed out for the use of more efficient building floors solutions, characterized by a good structural response combined with lightness, slenderness, flexibility, and economy, when compared to conventional approaches. The development of structural systems with these characteristics frequently results in floors with low damping ratios and natural frequencies values close to those associated with human activities, such as walking, running, jumping, and dancing, inducing resonance, excessive vibration problems, and human discomfort. This way, in this study, biodynamic models are utilised to represent the effects of rhythmic human actions, specifically jumps performed in aerobics classes, on reinforced concrete building floors. To do this, Artificial Neural Networks (ANNs) were developed to generate biodynamic models associated with single degree of freedom (SDOF) mass-spring-damper systems. Based on these biodynamic models, it becomes possible to evaluate the dynamic structural response of building floors when subjected to rhythmic jumping. The proposed analysis methodology aims to demonstrate that the models developed from ANNs are capable of considering the dynamic contributions that individuals exert on building floors structures, standing out as a viable and more realistic alternative in comparison with the traditional mathematical models of dynamic loading (only force models).

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13674 2025-12-18T20:47:03+00:00 Giovane Avancini giovanea@unicamp.br Nathan Shauer shauer@unicamp.br Gustavo Henrique Siqueira siqueira@fec.unicamp.br Philippe Remy Bernard Devloo phil@unicamp.br

This work presents a locally conservative scheme to solve two-phase flow problems in heterogeneous porous media. The Darcy equations are discretized using a mixed formulation of the Finite Element Method (FEM) based on de Rham compatible spaces; i.e. H(div, Ω) functions to approximate the flux field and L2(Ω) discontinuous functions for the pressure field. These choices ensure that mass conservation holds strongly and provide highorder accuracy for the flux variables. For the transport problem, piecewise constant functions (Finite Volume Method) are used to approximate the saturation field. Quadratic relative permeability curves with residual saturation is adopted for both water and air, and a linear density function is considered to account for the air compressibility. The time-marching procedure is performed using a standard Backward-Euler method. The nonlinearity due to the coupling is solved using a Sequential Fully Implicit (FSI) strategy. The proposed scheme is applied to simulate an experiment on filling an initially unsaturated experimental apparatus using different water injection procedures.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13633 2025-12-18T20:45:37+00:00 Gabriele Lins gabriele.lins@ctec.ufal.br Otávio Bruno de Araújo Rodrigues otavio.rodrigues@lccv.ufal.br Catarina Nogueira de Araújo Fernandes catarina@lccv.ufal.br Thiago Barbosa da Silva thiago.barbosa@lccv.ufal.br William Wagner Matos Lira william@lccv.ufal.br

This work proposes to analyze and implement analytical formulations for the calculation of axial loads with buckling in tubing strings in oil wells. These strings are subjected to different operational conditions throughout the well’s lifespan. Accurately predicting the axial loads resulting from these operations is essential to prevent accidents and ensure structural integrity. Such loads arise from pressure and temperature variations, self-weight, and buckling—the latter being the main focus of this work. To achieve the proposed objective, the adopted methodology is divided into six main stages: (i) literature review on oil wells; (ii) study of analytical solutions for axial load prediction in tubular strings; (iii) investigation and inclusion of buckling effects in analytical equations; (iv) definition of a vertical well configuration for implementation of the studied equations; (v) Well modeling using the SCORE tool; and (vi) comparative analysis of the obtained results. The analytical solution showed excellent agreement with the reference numerical tool, allowing for the quantification of the buckled length and its impact on the axial forces. The next step of this work involves the numerical modeling of buckling and axial loads in tubing strings, contributing to advancements safe design, especially in environments with complex geometries and more severe operational conditions.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13649 2025-12-18T20:46:14+00:00 Moisés dos Santos moises.santos16512@gmail.com Renato Ludwig Pilan renatopilan@yahoo.com.br Leonardo Avane leonardoavane@gmail.com Hugo Sakai Idagawa hugo.sakai@sp.senai.br Vitor Frutuoso de Carvalho vitor.frutuoso@senaisp.edu.br Leonardo Gabriel Siveri Leonardosiveri@yahoo.com.br

Modal analysis is a technique used to identify the dynamic properties of structures, such as natural frequencies, vibration modes and damping. This analysis is essential to predict structural behavior under different excitation conditions and to ensure the integrity and safety of projects. Among the instruments that perform modal tests, the impact hammer is generally used, however the commercial models available on the market are expensive, which makes them difficult for educational institutions with limited budget to access. Thus, this work developed an impact hammer designed to measure the excitation impact force in small to medium-sized structures which will also perform modal analysis. The objective was to create an economically viable and more accessible alternative when compared to the equipment available on the market. Furthermore, the project serves as a didactic kit, facilitating the teaching and understanding of the principles related to structural vibration. In order to perform the appropriate measurements, the hammer has a load cell, seismic mass and analog electronics, this allows the measurement of impact forces and perform the frequency response analysis of a body. Additionally, this instrument was developed through machining, which allowed the creation of an internal cavity to house the components. The handle was produced using stainless stell, designed to internally incorporate the electronic circuits for the system to operate. Calibration was performed using Bump Test and based on Newton's Second Law, using the previously determined seismic mass and the acceleration of gravity as a reference. To this end, a device was built that standardizes the acceleration applied to the system, allowing repetitions and controlled impacts. Moreover, in each test, the electrical voltage generated by the compression of the piezoelectric cell was analyzed. The final validation of the system was an experiment in which a clamped steel bar was subjected to impact excitation. The results achieved were satisfactory when compared with the theoretical model of the bar and with data obtained from a commercial instrument, ensuring accuracy with an error of 5%.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13665 2025-12-18T20:46:46+00:00 Murilo Henrique Campana Bento m.bento@usp.br Sergio P. B. Proença persival@sc.usp.br C. Armando Duarte caduarte@illinois.edu Nathan Shauer shauer@unicamp.br

The Generalized/eXtended Finite Element Method (G/XFEM) is a numerical method whose approximation space is obtained by augmenting standard FEM spaces with functions designed to capture well the behavior of local features of the problem under investigation, such as cracks and material interfaces. In the context of linear elastic fracture mechanics (LEFM) problems, the G/XFEM significantly alleviates mesh requirements and facilitates mesh generation by allowing cracks to cut through finite elements, with the discontinuity inside an element being captured by enrichment functions. At the same time, it also enables using coarser meshes close to crack fronts if the appropriate singular enrichments are adopted. In the case of second-order G/XFEM formulations for 3-D LEFM problems, mesh refinement close to crack fronts is still needed to reach optimal convergence rates if singular enrichment functions that represent only the square root of r singularity are adopted. To address this, we present a computationally efficient and robust h-adaptive algorithm. The procedure is driven by a block-diagonal Zienkiewicz and Zhu (ZZ-BD) error estimator, specifically tailored to estimate well discretization errors of 3-D LEFM problems solved by second-order G/XFEM. It also enables automatic non-uniform mesh refinement around crack fronts, effectively resolving problems with strong three-dimensional effects. The proposed h-adaptive strategy recovers optimal convergence rates for second-order G/XFEM and holds practical significance, as it delivers final discretizations on the fly that satisfy a user-defined tolerance for the discretization error. The influence of this h-adaptive technique on quantities of interest, such as stress intensity factors, is also investigated in this contribution.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13624 2025-12-18T20:45:15+00:00 Gabriela Pereira Lubke Izoton gabriela.lubke@ifes.edu.br Fernanda dos Santos Franco fernandafranco021@gmail.com Sidineidy Izoton sidineidy.izoton@hotmail.com

Structural optimization has emerged as an essential tool in engineering, driven by the growing demand for more conscious and efficient use of resources. This context supports the adoption of computational optimization techniques in civil construction projects, especially in the development of lighter structural elements with improved mechanical performance. Silva and Lübke[1] applied the finite element software FEMOOP, developed in C++, for the modeling and analysis of steel open web beams. The software includes the formulation of a three-node triangular finite element under plane stress conditions, along with specific routines for shape optimization and the detailing of the ideal cutting line based on the obtained model. The present study aims to evaluate the influence of varying parameters used in genetic algorithms on the results produced by the implemented optimization routine. Different parameter combinations—such as initial population size, crossover rate, mutation rate, and elitism—were tested and applied to the optimization of different open web beams. The results indicate that proper calibration of these parameters can significantly affect the algorithm’s performance and the quality of the solutions generated, directly impacting the load-bearing capacity of the optimized beams. Thus, this study contributes to the advancement of the structural optimization process, providing guidance for selecting more efficient configurations when applying genetic algorithms to the design of web open steel beams.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13640 2025-12-18T20:45:53+00:00 Carlos Freire dos Santos carlos.freire.oil@gmail.com Paulo Cesar Philippi paulophilippi@gmail.com

Porous structures are essential in nature and particularly relevant in petroleum reservoirs, as they allow for the storage and migration of fluids such as oil, water, and gas. Understanding these properties is crucial for optimizing exploration and reducing operational costs. Computational simulation methods, such as the Lattice Boltzmann Method (LBM), provide an efficient way to model these phenomena and are especially effective in porous media with complex geometries, supporting studies of multiphase flows.The color gradient method based on LBM is an efficient tool for simulating two-phase flows, allowing independent control of surface tension and interface thickness. It stands out for its ability to accurately represent stationary bubbles and phase coalescence, making it ideal for studying bubble stability and dynamics in porous or confined media.The article proposes two improvements to the color gradient model in the Lattice Boltzmann Method, aiming for greater stability and accuracy in simulations of immiscible displacement in narrow porous media. The first is the introduction of a volume exclusion term, which reduces compressibility and prevents numerical instabilities. The second is the adoption of the TRT (Two-Relaxation-Time) model, which eliminates second-order Knudsen number effects and improves physical consistency. Modifications to the segregation rules and surface force formulations are also suggested, making the model more robust and suitable for complex geometries.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13656 2025-12-18T20:46:29+00:00 Kaique Magalhães kaique.magalhaes@ufba.br Renato de Paula S. Trindade trindadeestruturas@gmail.com Elder Alves Satos elderbx@hotmail.com Reyolando M. L. R. F. Brasil reyolando.brasil@usp.br

This study investigates the phenomenon of stiffness enhancement in structures used in aerospace vehicle components subjected to geometric nonlinearity. Geometric nonlinearity occurs when the deformation of a structure results in significant changes to its geometry, leading to alterations in stiffness. Understanding this behavior is crucial for the design and analysis of structures in various aerospace engineering applications. Through numerical simulations, this study investigated the mechanisms behind stiffness enhancement by analyzing its natural frequency and its implications on structural performance. Ultimately, an increase in stiffness and, consequently, in the natural frequency of internally pressurized structures was observed.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13672 2025-12-18T20:46:59+00:00 Gustavo Almeida de Moura gustavo.almeeida92@gmail.com Paula de Oliveira Ribeiro paula.oliveira@ufjf.br Fernanda Giannotti da Silva Ferreira fgiannotti@ufscar.br Pablo Augusto Krahl pabloaugustokrahl@gmail.com Silvete Mari Soares silvetemari@ifsp.edu.br José Américo Salvador Filho jasalvador@ifsp.edu.br

This study addresses the challenge of predicting chloride ingress in UHPC and depassivation of reinforcing steel, which is essential for durability assessment. A numerical model was calibrated using experimental data to estimate the time for chloride to reach the critical threshold (0.3%) at reinforcement depth. Simulations considered cover depths of 15, 20, and 25 mm under chloride concentrations of 0.5% and 1.95%. Results show that under low chloride exposure, the UHPC cover can protect steel for more than 100 years before depassivation, whereas under severe marine conditions, this duration drops to 16–25 years, depending on the cover. Matrix diffusivity and chloride concentration are the key parameters governing chloride ingress and the time to reach 0.3% at reinforcement depth, critically influencing service-life prediction. Numerical simulation proved efficient in overcoming the time limitations of experimental programs and supports service-life design.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13631 2025-12-18T20:45:30+00:00 Flavio Canfilde Alves Pereira flaviocanfilde@ufpr.br William Rodrigo Lüdtke william.ludtke@ufpr.br Giuliana Sardi Venter giuliana.venter@ufpr.br João Vitor de Carvalho Fontes fontes.joao@ufpr.br

Stewart-Gough Platforms (SGPs) are parallel manipulators with various applications, such as wave stabilization of off-shore structures in the oil and gas industry, due to it’s precise position and orientation control across all six degrees of freedom. Associating an SGP with a ball-and-plate system (BPS), where a ball is allowed to freely roll over an actuated plate, can provide a test bed for nonlinear trajectory control strategies. This study aimed to develop an analytical simulation model of the behavior of a physical SGP-BPS system, and to validate it using empirical data. The model was based on principles of analytical geometry and kinematics, with the objective of accurately replicating the behavior of a physical platform, available at the Laboratory of Monitoring and Control from the Federal University of Paraná (UFPR). This platform was actuated by six servomotors, whose angular positions determined the attitude of a plate equipped with a resistive touchscreen sensor. A free-rolling steel sphere was placed on top of the plate and had its coordinates continuously read by a micro-controller that determined and fed the necessary servo angles to the system to adjust the sphere position and/or trajectory as intended. The sphere position was recorded over time, as well as timestamps and individual angles for each actuator, during several experiments covering a wide range of operating scenarios. The analytical model was assessed using the data collected during the operation of the physical platform. The model was provided with the initial position and velocity of the sphere, as well as the angular positions of each actuator during the operation of the platform. This stream of angular inputs and timestamps allowed the model to predict the plate orientation and the sphere position and velocity over time. The model accuracy was evaluated by comparing the sphere’s simulated position and velocity with the experimental data, for the same set of input angle values. The main parameters for this evaluation were the euclidean norm error and absolute maximum error between the simulated position and the actual position of the ball over time. Results showed a good correlation between model predictions and experimental data. The analytical model is expected to be applied on future works in the creation of a digital twin of the platform for real-time control applications, and the development of a data-driven controller for off-shore wave stabilization systems.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13647 2025-12-18T20:46:09+00:00 Arthur Gomes arthurbg951@hotmail.com Roberto Quevedo rquevedo@tecgraf.puc-rio.br Deane Roehl deane@tecgraf.puc-rio.br Bruno R. B. M. Carvalho brcarvalho@petrobras.com.br

The definition of damage zones in geological faults is important for oil and gas exploration due to the potential occurrence of preferential flow paths through such zones. In practice, some damage zone thicknesses in isolated faults are evaluated using exponential laws based on statistical relationships with fault displacement. However, in the field, linking damage zones resultant from fault interactions are observed, compromising the production process in reservoirs. One way to estimate the geometry of linking damage zones is through numerical modeling using the Finite Element Method and elastoplastic constitutive models. Those models impose displacements or throws over fault surfaces, triggering plastic zones obtained in Gauss Points around the faults which are related to the damage zones. However, model building, simulation execution, and result analyses can be time-consuming, hindering decision-making in field operations. As an alternative, the present study proposes the construction of a neural network to predict damage zones resulting from the interaction of two geological zones. The dataset used was built from numerical results of several analyses using a set of geometrical parameters such as overlap, separation, and angle between two faults. This strategy is interesting because it reduces the dimensionality of the number of parameters to be sampled while maintaining the key information to reconstruct the data in global coordinates. The fault setting as well as the results obtained from the simulations at each Gauss point were used as input parameters to train the neural network in order to identify the damage zones given by plastic deformations. A regression strategy is used to preserve the magnitude of the results. The Adaptive Moment Estimation (ADAM) and hyperbolic tangent (Tanh) is being used as optimizer and activation function, respectively. 80% of the dataset is being used for training and 20% for testing, resulting in approximately 9 million data points for adjustment and 2 million for validation. The preprocessing used for network inference consists of generating a grid of points considering a fault arrangement. After inference, the network predicts the equivalent plastic strain field. In this way, it is possible to interpret the results obtained using contour lines. Therefore, it is expected to predict the occurrence of linking damage zones more quickly while considering the geometry of two faults.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13663 2025-12-18T20:46:42+00:00 Gabriel Gaynett Leturiondo gabrielgleturiondo07@gmail.com Rogério José Marczak rato@mecanica.ufrgs.br Felipe Tempel Stumpf felipe.stumpf@ufrgs.br

One of the most commonly used criterion to determine the structural failure of subsea equipment is found in API standard 17G, according to which the collapse takes place if, for any individual component or part, the average equivalent plastic strain in the through section path reaches 2% in the deformed configuration. This is known as "the 2 % strain method". In the present work we investigate alternative methodologies to determine the structural failure of an equipment using approaches based on strain energy. We conduct a series of numerical simulations of structures with prior damaged geometries and determine their structural capacity based on the 2% strain method of API 17G. Those results are used as input for posterior analyses in which the strain energy around the damage is measured (numerically) so we can correlate the ultimate load and the energy absorbed in the vicinity of the damaged area. The results are used for successfully predicting the failure load of different geometries with different damaged areas and a critical discussion is carried out concerning the conservativeness of the API 17G criterion.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13679 2025-12-18T20:47:19+00:00 Sofia Nascimento da Silva sofianascimento307@gmail.com Erick Miguel Barbosa dos Santos erickmiguelbsantos@gmail.com Katarina Veljovic katarinaveljovic123@gmail.com Karin Komati profkarin@gmail.com

The rise of digitally manipulated audio content creates new challenges in verifying information authenticity, especially on social media. Advances in artificial intelligence (AI), particularly in text-to-speech and voice synthesis technologies, have greatly enhanced the quality and realism of generated audio. This study addresses the problem of deepfake audio detection and offers two main contributions. First, it introduces the FakeBrAccent dataset, which includes 746 audio samples (373 real and 373 synthetic) in Brazilian Portuguese, featuring regional accents such as Baiano (Bahia), Fluminense (Rio de Janeiro and Espírito Santo), Sulista (Southern Brazil), Nordestino (Northeastern Brazil), and Carioca (Rio de Janeiro city). The original BrAccent dataset was used as both the source of real samples and as a reference for simulating accents during the generation of synthetic samples with a text-to-speech tool. Second, the study evaluates the performance of two classification models, Convolutional Neural Networks and XGBoost, on this dataset. The models were tested using standard performance metrics, including accuracy, precision, recall, and F1-score. The findings provide a baseline for future research into synthetic speech detection in Brazilian Portuguese, emphasizing the role of accent variation in model performance.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13622 2025-12-18T20:45:11+00:00 Marcello Congro marcellocongro@puc-rio.br Pedro A. L. P. Firme plobo@tecgraf.puc-rio.br Deane Roehl droehl@puc-rio.br

Ensuring long-term well integrity remains a critical challenge in plug and abandonment (P&A) and carbon capture, utilization, and storage (CCS/CCUS) operations, particularly in clayey formations like shales that exhibit time-dependent deformation. In such environments, wellbore closure due to creep forming a geological barrier contributes to zonal isolation and long-term sealing. This study proposes a modified creep analytical model that extends the classical Barker solution, originally developed for salt formations, to capture the creep behavior of shales. The formulation adopts deviatoric stress instead of the difference between horizontal stress and wellbore pressure as the main driver of time-dependent deformation, consistent with viscoelastic/viscoplastic constitutive models relevant to salt and shale geomechanics. The stress state surrounding the wellbore right after drilling is derived using Kirsch’s elastic solution, enabling a closed-form expression of the deviatoric stress field. Next, creep behavior is activated, and a time-dependent exponential decay function, calibrated using finite element simulations in ABAQUS, is introduced to capture progressive stress relaxation during material deformation. The resulting stress field is incorporated into a semi-analytical, iterative framework for computing wellbore radius evolution over time, based on a creep law that depends on stress, temperature, and empirically defined material parameters. The model is implemented in Python within a lightweight computational framework, offering rapid, yet robust, assessments suitable for field-scale applications. Validation against numerical simulations demonstrates strong agreement, underscoring the model’s applicability for integrity evaluation and barrier design in P&A and CCS/CCUS scenarios. The proposed framework offers a computationally efficient, physically consistent approach for predicting long-term wellbore performance in creeping shale formations.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13638 2025-12-18T20:45:49+00:00 Saul Aluce saulaluce10@gmail.com Guilherme Santana Alencar guilherme.alencar@unb.br Lenildo Santos da Silva lenildo@unb.br Hermano Cardoso hermano.cardoso@arcelormittal.com.br Débora Maia Guimarães debora.m.guimaraes@arcelormittal.com.br

This contribution presents a web-based structural application for the design and verification of composite steel-concrete slabs, based on strength, stability, and serviceability limit state criteria. Its intuitive interface, accessible via browser on both desktop and mobile devices, enables engineers to efficiently perform structural calculations, including the determination of design actions and resisting capacities, as well as the verification of code-based requirements, shrinkage reinforcement, and, when applicable, negative moment reinforcement. The system also includes structural fire design checks and dynamic performance evaluation related to vibration serviceability. The application is structured into three core modules: a user interface (frontend), a data management layer (backend), and a processing module responsible for executing structural simulations. Once the input data is submitted, the backend processes the information and forwards the parameters to the analysis engine. Results are stored and presented to the user in a clear and organized format. The accuracy of the results has been validated against theoretical formulations and recent updated design standards, including NBR 8800:2024, AISI S100:2024, and NBR 14323:2013. The results obtained confirm the tool’s applicability to practical structural design scenarios. This solution contributes to the modernization of composite slab design in civil engineering, enhancing calculation efficiency, precision, and structural analysis automation.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13654 2025-12-18T20:46:25+00:00 Thyago Hannan th.hannan.19@gmail.com Guilherme Santana Alencar guilherme.alencar@unb.br José Guilherme Santos da Silva jgss@uerj.br

The Honestino Guimarães bridge was inaugurated in the mid-1970s and is the only structure of its kind in Brasília/DF, Brazil, designed by the renowned architect Oscar Niemeyer, where his understanding of structural systems was used aiming to create a symbolic and harmonious environment, having in mind a structural system composed by steel and concrete. The central span of the bridge is made of steel, with an orthotropic deck, presenting an arched shape, Gerber joint connection, and longitudinal symmetry. This way, this research work aims to investigate the non-deterministic dynamic structural behaviour and assess the structural fatigue on the central span of the bridge. A global finite element model and a numerical submodel were developed using the ANSYS software, incorporating the main physical and mechanical characteristics of the bridge’s central span, which measures 58 m in length and 13.50 m in width, with the concrete slab thickness varying from 0.07 m to 0.12 m. Thus, the hot-spot stresses were determined from the submodel created from shell elements (SHELL188), and identified through the use of the Vehicle Bridge Interaction (VBI) computational software, which integrates the ANSYS and MATLAB software. It was observed that, as the vehicle velocity increases, the time interval in which hot-spot stresses reach the maximum values decreases. The interval between maximum stresses shortens as the vehicle velocity increases, indicating faster and more concentrated dynamic loads. The intensity of hot-spot stresses is more influenced by the number of vibration modes than by the vehicle velocity, highlighting the importance of appropriate numerical modelling in order to capture the dynamic effects.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13670 2025-12-18T20:46:55+00:00 Pedro Felype faonsup@gmail.com Guilherme Silveira Da Rocha guisilveirarocha@gmail.com Juliette Zanetti juliette.zanetti@multivix.edu.br Yuri Ricardo Moratori yrmoratori@gmail.com

According to data from eSocial, part of the Ministry of Labor and Employment, nearly 500,000 work-related accidents were recorded in Brazil in 2023, 2,888 of which were fatal. A survey by the Occupational Health and Safety Observatory, covering data from 2012 to 2022, revealed that the country registered more than 7 million work accidents during that period, considering only formally employed workers under the CLT regime. Currently, Brazil ranks 4th in the global ranking of occupational accidents, behind only China, India, and Indonesia. These figures highlight the importance of occupational safety, especially in industrial and construction environments, where the proper use of Personal Protective Equipment (PPE), such as safety helmets, is essential to prevent accidents and protect workers' physical integrity. However, PPE compliance is still predominantly monitored manually, which poses significant limitations, particularly in areas with high worker traffic. In light of this scenario, this project proposes the development of an intelligent real-time monitoring system capable of automatically detecting the absence of safety helmets using computer vision techniques applied to images captured by surveillance cameras. To achieve this, the YOLO (You Only Look Once) convolutional neural network was employed, a model widely recognized for its high accuracy and efficiency in real-time object detection tasks. The developed system identifies workers not wearing helmets, captures images of the infraction, and automatically sends alerts to the company’s monitoring center. The methodology involved image collection, model training using YOLO, system integration development, and testing in simulated environments. The results showed an accuracy of over 90% in detecting the absence of helmets, demonstrating the effectiveness of the proposed solution. In addition to automating compliance monitoring, the system reduces reliance on human supervision, strengthens adherence to regulatory standards, and promotes safer workplaces. The proposal also stands out for its accessibility, robustness, and practical applicability, offering a promising alternative for modernizing occupational safety management.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13629 2025-12-18T20:45:24+00:00 JUAN FRANKS VALENZUELA CARRASCO 021200171j@uandina.edu.pe Elvis Yuri Mamani Vargas emamaniv@gmail.com Luis Fernando Paullo Muñoz lfernand@tecgraf.puc-rio.br

This study aims to evaluate the structural behavior of a suspension bridge composed of natural fiber cables. For this, a structural-computational analysis of the Q’eswachaka bridge is performed. The Q’eswachaka bridge is an Inca heritage structure that is rebuilt annually using Q’oya fibers (Festuca dolichophylla). A numerical model was developed in the structural analysis software Abaqus/CAE using a finite element mesh composed of truss elements. The mechanical equilibrium under self-weight and pedestrian loads is assessed considering geometric nonlinearities, assuming linear-elastic material behavior. The model’s validation is carried out at two distinct levels. Firstly, the FEM model is compared with a simplified analytical solution based on the response of an inextensible catenary cable. Secondly, an external contrast is performed using the classical experiment by Irvine Sinclair (1976) for cables under off-center loads. The proposed framework also includes a structural sensitivity analysis considering variations in key structural parameters such as the elastic modulus (E), cross-sectional area (A), linear weight (w), and initial sag. In addition, this work aims to identify critical behavioral thresholds. The methodology enables the estimation of the structural working range of the bridge under different service-load scenarios, avoiding destructive testing and preserving the integrity of the heritage asset. This approach provides a replicable basis for the structural evaluation and conservation of traditional suspension bridges, integrating empirical knowledge with advanced computational modeling.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13645 2025-12-18T20:46:05+00:00 Sidnei Alves de Araújo saraujo2107@gmail.com Gustavo Araujo Lima gustavo.araujo.lima94@gmail.com Sergio Vicente Denser Pamboukian sergio.pamboukian@mackenzie.br Vitor Pessoa Colombo vitor_pessoacolombo@gmail.com

The use of drones in combating mosquito breeding sites, such as Aedes aegypti, is already a consolidated reality. However, although the presence of stagnant water is an essential condition for the reproduction of these vectors, computer vision approaches reported in the literature rarely address the direct detection of water in suspected objects and scenarios, representing a significant research gap. In this work, we propose a method based on Genetic Algorithms (GA) for creating a water index, obtained from arithmetic combinations of the image spectral bands. Each GA chromosome, when decoded, generates an index that is evaluated by a fitness function aimed at minimizing the difference between the pixel values produced by the application of the index and the expected values, extracted from manually annotated images. In summary, at the end of the evolutionary cycle, the GA produces a water index that, when applied to an image captured by the drone, generates a binary image highlighting the regions corresponding to the presence of water. The results obtained (accuracy = 87.2%, precision = 85.2%, and sensitivity = 95.2%), demonstrate the feasibility of the proposed method for detecting small water accumulations in objects and scenarios potentially associated with breeding sites, thus enhancing the effectiveness of drones in surveillance and epidemiological control activities.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13661 2025-12-18T20:46:38+00:00 Thiago Artur Mendes de Souza thiago.arthursouza@gmail.com Giedre Alves Sirilo giedresirilo@hotmail.com Túlio Nogueira Bittencourt tbitten@usp.br João Victor Fragoso Dias joao.v.dias@ufes.br Hermes Carvalho hermesc2000@gmail.com

Calibration techniques are essential for applying finite element models (FEM) in structural assessment and monitoring, given the uncertainties inherent to modeling. Among indirect calibration approaches, metaheuristic algorithms stand out for their ability to explore large search spaces and avoid local minima, but they require many costly model evaluations. To address these limitations, surrogate-based approaches have emerged, approximating the behavior of the objective function and substantially reducing the number of physical model evaluations needed to obtain suitable solutions. Bayesian Optimization (BO), for example, employs a probabilistic model to estimate the target function, enabling not only efficient optimization but also providing quantitative insights into parameter sensitivity and prediction uncertainty. In this work, the performance of a widely used metaheuristic, the Genetic Algorithm (GA), is compared with that of BO in the calibration of a synthetic bridge model, using modal data with added noise. The results demonstrate BO’s versatility, competitive accuracy under noisy conditions, and markedly lower computational cost, while also indicating that hybrid strategies can further improve precision.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13677 2025-12-18T20:47:13+00:00 Silvia Lorena Bejarano Bermudez silviabermudez@petroleo.ufrj.br Gabriel Sousa dos Santos Ribeiro gabriel.ribeiro@petroleo.ufrj.br Paulo Couto pcouto@petroleo.ufrj.br Thiago Pineiro thiago.pineiro@petroleo.ufrj.br José Luis Drummond Alves jalves@lamce.coppe.ufrj.br Maira da Costa de Oliveira Lima Santo maira.lima@petroleo.ufrj.br Austin Boyd austin@petroleo.ufrj.br Luciano Barros Guedes luciano.guedes@petroleo.ufrj.br

Free fluid porosity (FFI) is a key petrophysical parameter for reservoir characterization, representing the fraction of the porous volume that contains potentially producible fluids. It is commonly estimated from transverse relaxation time (T₂) distributions derived from Nuclear Magnetic Resonance (NMR), using a cutoff value to distinguish free (mobile) fluids from irreducible ones. The latter are retained within micropores due to capillary forces or adsorption onto mineral surfaces. Precise quantification of free fluid porosity is critical for estimating recoverable hydrocarbon volumes, particularly in hybrid carbonate reservoirs with mixed siliciclastic and carbonate facies, where standard NMR interpretations may be limited. This study investigates a novel approach based on digital erosion of segmented X-ray micro-computed tomography (µCT) images to estimate free fluid porosity in hybrid carbonate rocks, using coquinas from the Morro do Chaves Formation as a case study. In addition, the image-derived FFI was integrated with NMR measurements to assess its potential as an input for permeability estimation using the Timur-Coates model. The dataset includes six segmented pore space images from coquina samples that were previously centrifuged to determine irreducible water saturation (Swirr) and establish T₂ cutoff values. The digital erosion technique was applied using Avizo software, applying successive layers of fixed thickness to the segmented pore space. The goal was to determine an optimal erosion width that approximates the average mobile fluid layer, maximizing the correlation between µCT-derived porosity and NMR-based FFP. This estimated porosity was subsequently used as input to the Timur-Coates permeability model, and the resulting predictions were compared to absolute permeability values measured by gas flow on corresponding core plugs. The results confirmed the existence of an optimal erosion width that maximizes the correlation between µCT-derived and NMR-derived FFI. Furthermore, the Density Equation (DE) approach was applied to differentiate between macropores (associated with mobile fluids) and micropores (associated with bound fluids), providing an additional comparative framework alongside the digital erosion technique. This integrated method provides a robust, non-destructive workflow to estimate FFI, improving petrophysical characterization and modeling of complex porous media, especially in hybrid carbonate reservoirs.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13620 2025-12-18T20:45:05+00:00 Alana Camata Carpanedo camatacarpanedoalana@gmail.com Gustavo Alves Lima 2001gustavoalves@gmail.com Werley Gomes Facco werleyfacco@ifes.edu.br

Electric motors play a central role in the automation and operation of modern industrial systems and are widely used in production processes around the world. These devices have the primary function of converting electrical energy into mechanical energy through the interaction between magnetic fields and conductors carrying electric current. Among the various available topologies, the three-phase induction motor with squirrel-cage rotor stands out for its robustness, low maintenance cost, and wide applicability. It is the most widely used in industry due to its ability to efficiently meet the operational demands of various sectors. This work aims to study relevant electromagnetic quantities, such as the electromagnetic torque in the squirrel-cage rotor motor. For this study, the Finite Element Method implemented in MATLAB will be used to simulate and evaluate relevant electromagnetic quantities, based on the geometric and electrical parameters of the Tesla Model S electric motor, available in the literature. The electromagnetic behavior of the induction motor with squirrel-cage rotor—whose characteristic structure includes short-circuited bars—was simulated. The motor parameters were incorporated into the developed algorithm, allowing for the analysis of the distribution of the magnetic vector potential and the flux in the air gap. Time harmonization was taken into account, ensuring greater ease in modeling and contributing to the achievement of results that proved to be consistent with data from the literature. The extended paper will present the details of the computational modeling, the FEM formulation applied to the electromagnetic problem, as well as graphs, field distributions, and quantitative comparisons between the simulated results and reference values.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13636 2025-12-18T20:45:46+00:00 Fabricio Simeoni de Sousa fsimeoni@icmc.usp.br Uebert Gonçalves Moreira uebert.moreira@usp.br Franciane Fracalossi Rocha fr.franciane@alumini.usp.br

Karst conduit systems exhibit complex geometries, where networks of channels, cavities, or vugs coexist within a porous matrix. Accurately simulating fluid flow in such heterogeneous environments remains a challenge that often limits the reliability of predictions. This study presents a conservative finite volume framework that ensures mass conservation while coupling the karst conduit and porous matrix domains to simulate coupled 1D/3D flows in carbonate rocks. The method is designed to handle intricate networks of karst conduits, including cases in which individual branches have distinct properties—particularly differing permeabilities and size ratios. The karst subdomain is embedded within the matrix grid to ensure geometric consistency and alignment, while mass exchange between the two domains is governed by a transmissibility factor. Numerical experiments include a sensitivity analysis of key parameters, such as permeability contrast and branch ratio, to examine their effects on pressure and velocity fields and to assess the model’s ability to capture these variations.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13652 2025-12-18T20:46:21+00:00 Silvio Ney Alves Veras pg405394@uem.br Daniel William Costa de Avelar pg56009@uem.br Ana Caroline Caetano pg405997@uem.br Wilson Wesley Wutzow wwwutzow@uem.br

The accurate assessment of structural elements with local geometric discontinuities, such as notches, remains a recurrent challenge in civil engineering, especially under dynamic loading conditions. Notches introduced due to architectural, constructional, or functional requirements act as critical regions of stress concentration, potentially compromising structural integrity and triggering instabilities even in otherwise robust systems. In this context, the use of models capable of accurately capturing geometric nonlinear effects becomes essential, particularly in scenarios involving large displacements.This study presents a numerical analysis of the dynamic response of two-dimensional notched beams using the Positional Finite Element Method (FEM-P), in which the equilibrium equations are formulated in the deformed configuration. Unlike conventional formulations based on nodal displacements, FEM-P adopts nodal positions as primary variables, allowing a more direct treatment of geometric nonlinearity, especially in simulations involving large displacements. Time integration is performed using the implicit Newmark-beta scheme, ensuring numerical stability and good accuracy. Additionally, the adopted constitutive model is the Saint Venant–Kirchhoff hyperelastic formulation, suitable for large elastic deformation regimes.The computational code developed was initially verified through classical benchmark examples widely referenced in the literature on solid dynamics, aiming to ensure the fidelity of the numerical implementation against established solutions. After this verification stage, the model was applied to the analysis of notched beams with various geometric configurations, including variations in depth, shape, and location of the notches, in order to investigate their effects on internal stress redistribution and overall dynamic response.The results indicate that the presence of notches significantly contributes to displacement amplification, redistribution of internal forces, and early onset of geometric instabilities. The proposed approach highlights the importance of advanced modeling techniques for structures with geometric discontinuities, especially in scenarios where the interaction between nonlinear kinematics and dynamic excitations may compromise structural integrity.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13668 2025-12-18T20:46:51+00:00 Felipe Pamplona Mariano fpmariano@ufg.br Gino Bertollucci Colherinhas gino@ufg.br Paulo Henrique Neves Pimenta engmecpaulohenrique@gmail.com Gabriela Miguel Fraga gabriela.fraga@goias.gov.br Getulio Santiago dos Santos Junior getulio.santos@goias.gov.br

This paper presents the development of a digital platform for managing and assessing energy performance in public buildings, with a pilot application in the state of Goiás, Brazil. The system, implemented in Python using the Flet framework, integrates three interconnected SQLite databases covering electricity billing, certified equipment performance data from Brazil’s National Energy Labeling Program (PBE-INMETRO) and on-site audit records. It enables the automated import and structured extraction of key billing information such as consumption, demand, tariff group and cost from PDF and Excel files, as well as the detailed registration of installed equipment at the room level. All entries are cross-referenced with standardized efficiency data to ensure consistency and accuracy. The platform provides interactive dashboards for consumption analysis, supports both macro-level and micro-level diagnostics and facilitates transparent monitoring of energy use patterns over time. A pilot case study was carried out at the Secretaria-Geral de Governo de Goiás (Goiás State Government’s General Secretariat – SGG), where the tool replaced manual paper-based data collection with direct digital registration, significantly improving data acquisition efficiency. The results highlight the platform’s potential to inform public policy, guide targeted energy efficiency interventions and serve as a replicable and open-access solution for other public institutions. Its structured architecture also establishes a foundation for the analysis of large data repositories through advanced data analytics techniques, enabling large-scale pattern recognition, trend forecasting and enhanced decision-making capabilities. Future developments include advanced dashboard features for year-to-year comparisons, group-based consumption analysis and the integration of performance indicators to evaluate energy efficiency and end-use consumption shares by equipment category.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13627 2025-12-18T20:45:20+00:00 Nuno Pires nuno.pires@ctec.ufal.br Jennifer Mikaella Ferreira Melo jennifer.ferreira@lccv.ufal.br Beatriz Ramos Barboza beatriz@lccv.ufal.br João Paulo Lima Santos jpls@lccv.ufal.br Eduardo Toledo de Lima Junior limajunior@lccv.ufal.br Fábio Sawada Cutrim fabiosawada@petrobras.com.br Rafael Dias rafael_dias@petrobras.com.br Mávyla Sandreya Correia Tenório mavyla.tenorio@lccv.ufal.br

In computational simulations, time is a critical resource that often limits the scope and depth of analyses. Complex geotechnical simulations such as the hammer driven installation of conductor casings in oil‐well projects require extensive computational resources and frequent manual interventions during execution. This work presents the development of an automated agent that optimizes the simulation process for conductor casing installation using an open‐source software, a Material Point Method (MPM) simulator. The conventional hammer‐installation workflow in this tool exhibits significant limitations, as each hammer impact must be performed manually by the operator through successive modifications of input files at each simulation step. Considering that typical simulations involve more than 3,000 impacts, with an average processing time of 20 minutes per impact, manual intervention becomes impractical, leading to extended total simulation times and a high likelihood of operational errors.The Python based agent employs graphical interface automation techniques to manipulate input files and control the software’s execution, carrying out a three‐step cycle for each hammer blow: impact force application, force removal, and inter impact pause. The system automatically detects the current simulation stage, updates the required parameters in the control files, and issues the calculation commands entirely eliminating the need for human intervention. The soil is modeled as an undrained, cohesionless material using the Mohr Coulomb criterion, while the conductor casing is treated as a rigid body. Results show that automation delivers significant efficiency gains by reducing total simulation time and eliminating operational errors associated with manual handling. The agent’s implementation enables uninterrupted execution of hundreds of hammering cycles, allowing for more comprehensive and accurate analyses of soil behavior during conductor casing installation.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13643 2025-12-18T20:45:59+00:00 Philipe Queiroz Rodrigues engcivil.philip@gmail.com Eduardo Marques Vieira Pereira eduardo.marquesvp@gmail.com Rodrigo Bezerra Andrade rodrigo31ba@gmail.com Fábio Fleming Leitão fabiofleming@gmail.com Gustavo Henrique Siqueira ghsiq02@unicamp.br Hugo Luiz Oliveira hluiz@unicamp.br

Visual inspection techniques for structural damage detection are inherently limited: damage can be concealed, too minor for visual detection, or situated in inaccessible areas such as the interior of a nuclear reactor. To overcome these challenges, model-based approaches grounded in the solution of inverse problems have gained prominence. In particular, finite element model updating using static or dynamic response data has emerged as a powerful strategy to identify and quantify damage. By calibrating numerical models to reflect observed structural behavior more accurately, these methods support more informed maintenance planning and targeted intervention. This study presents a numerical strategy for identifying the structural parameters of a steel plate by minimizing a modified Constitutive Relation Error (MCRE) functional. Starting from a reference measure, the method seeks admissible mechanical fields and localizes regions with high errors. Once this iterative step converges, parameter updating is initiated to estimate the material properties accurately. The MCRE framework is based on the partitioning of the governing equations into a reliable set—where all equations are strictly satisfied—and an unreliable set, which encompasses potential deviations in the constitutive law. In the first phase, the admissible fields are obtained solving the global balance equations using a quadratic energy function that respects both Neumann's and Dirichlet's boundary conditions. In the second phase, the material parameters are identified through the minimization of a cost function that incorporates the previously obtained admissible fields and the reference measures. The CASTEM open-source toolbox for finite element computation was chosen to implement MCRE algorithm. The results obtained in this study demonstrate that the proposed numerical strategy was effective in accurately identifying both the stiffness parameters and the location of the defect in the plate.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13659 2025-12-18T20:46:34+00:00 Leonardo Wendler Felchak leonardo.wendler@ufpr.br Roberto Dalledone Machado rdm@ufpr.br Henrique Machado Kroetz henrique.kroetz@ufpr.br

Pultruded Fiber-Reinforced Polymers (FRP) profiles are innovative structural elements with outstanding mechanical properties, combined with economic viability. As some interesting properties of that material, it is possible to mention: its corrosion resistance, thermal insulation, high impact resilience, superior strength to-weight when compared to conventional civil industry materials like steel and aluminum, and cost-effective manufacturing processes. Nevertheless, studies regarding the effects of the heterogeneity of elastic and mechanical properties, due to the pultruded FRP profiles manufacturing process, on the reliability of those structural members under common loading conditions are scarce on the literature. So, in this sense the present work compares reliability analysis results of global buckling ultimate limit state of I-section pultruded beams, considering fiber-resin content uncertainty under homogeneous distribution and by considering the spatial variability of these properties using gaussian random fields. First Order Reliability Analysis (FORM) method and Monte Carlo Simulation (MCS) with Latin Hypercube Sampling were employed to calculate failure probabilities of pultruded FRP profiles under global buckling limit state. Buckling loads estimates were obtained by eigenvalue buckling finite element analysis. In order to reduce computational cost of the FEM models gaussian process metamodels were employed.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13675 2025-12-18T20:47:08+00:00 Bruno Bianchi Marques brunobm@usp.br Fernanda Gabriella Batista Santos Oliveira fernanda.gabriella@usp.br Ana Lucia Homce de Cresce El Debs analucia@sc.usp.br

The construction industry is a significant generator of solid waste, and the use of recycled aggregate concrete (RAC) mitigates environmental impacts. However, its application in partially encased composite (PEC) columns remains poorly explored. This study aims to simulate the structural behavior of PEC columns with RAC, considering coarse aggregate replacement ratios of 0%, 30%, and 50%, to evaluate the impact on their load capacity and axial strain response. The ABAQUS® numerical model was calibrated through experimental results and subsequently extended to include an additional replacement ratio for comparative purposes. The results indicate that the reduction in the mechanical properties of the concrete, associated with the gradual increase in the replacement ratio, resulted in a drop in both the load capacity and stiffness of the elements. Nevertheless, the overall structural response of the PEC columns remained consistent. These findings demonstrate the potential of RAC, showing that partial replacement of natural aggregates with recycled aggregates at low and intermediate substitution ratios maintains satisfactory structural behavior.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13634 2025-12-18T20:45:40+00:00 ígor Peter igorpeter@hotmail.com Alexandre Luis Braun alexandre.braun@ufrgs.br Felipe Schaedler de Almeida felipe.almeida@ufrgs.br

Billboards are a common feature in urban and highway environments, but their structural reliability under wind loads remains a concern. Many structural failures of billboard structures are observed every year, usually during weather events with high-intensity winds. Research indicates that recommendations found in standards around the world lead to underestimated wind load values for billboard structures, raising interest in further investigations in this field. A numerical investigation of steel billboards with a closed two-sided plate on flat terrain has been conducted with a numerical formulation based on finite elements and the explicit two-step Taylor-Galerkin model, used to solve the equations of fluid dynamics, considering incompressible and isothermal flow. The finite element meshes are made using linear hexahedral elements with reduced integration and hourglass control. Turbulence is handled with Large Eddy Simulation (LES) and a dynamic Smagorinsky subgrid model. The algorithm was implemented in CUDA-Fortran, utilizing the processing capabilities of the graphical processing unit (GPU) to speed up the simulation speed. The billboard model used corresponds to a full-scale billboard mounted at the Texas Tech University. Several meshes were constructed for different wind directions. The results are compared to full-scale measurements by other authors to validate the simulation. Mean and instantaneous velocity, and pressure fields, have been analyzed, alongside load and moment history. The GPU-enabled algorithm is compared with an equivalent CPU-only parallel algorithm, highlighting a speedup of up to 8.7 times compared to the CPU implementation.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13650 2025-12-18T20:46:16+00:00 Antonio Paulo Amancio Ferro antonio.ferro@lccv.ufal.br Erasmo Augusto Bezerra Silva erasmo.bezerra@lccv.ufal.br Francisco de Assis Viana Binas Júnior francisco.junior@lccv.ufal.br Lucas Pereira de Gouveia lucasgouveia@lccv.ufal.br Aline da Silva Ramos Barboza aline@lccv.ufal.br Aliel Faria Riente aliel@petrobras.com.br

Recent technological advancements in data transmission and acquisition systems for drilling operations, along with the growing global energy demand, have led the Oil & Gas industry to pursue data-driven solutions in real-time to improve performance and reduce well construction costs. In this context, the Rate of Penetration (ROP) serves as a key drilling performance metric, reflecting the effective speed of the drill string as the bit penetrates the rock formation. Predictive models for ROP can be employed in real time to suggest appropriate operational parameters that optimize ROP, potentially reducing drilling time and overall operating costs. This work investigates the application of Long Short-Term Memory (LSTM) networks for ROP prediction within a continual learning scenario based on drilling data. LSTM networks are well-suited for this task due to their ability to model dependencies in sequential data. The results obtained using the LSTM model are compared with benchmark results from Multilayer Perceptron (MLP) networks. A scenario is simulated in which the models are incrementally trained with new data received during drilling, predicting ROP in subsequent intervals. Given the transient nature of drilling, abrupt changes in data distribution are expected and may significantly impact predictive performance. Therefore, the model's performance in the test intervals is analyzed in relation to statistical characterizations of the newly received data. The dataset used comprises public data from three wells in the Volve field in the North Sea. It includes both operational parameters and lithological information, the latter being particularly important for analyzing the results obtained in this study. Model performance is evaluated using the Mean Absolute Error (MAE) and Root Mean Square Error (RMSE). The results demonstrate the potential of the adopted strategies using the LSTM model. It is expected that the comparative analyses presented in this study will contribute to key aspects of developing data-driven solutions for real-time drilling applications.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13666 2025-12-18T20:46:48+00:00 Kelson Pothin Wolff kelson.wolff@gmail.com Roque Luiz da Silva Pitangueira roque@dees.ufmg.br

Modeling three-dimensional crack propagation remains a challenging task in the context of computational methods, even for the widely used eXtended Finite Element Method/Generalized Finite Element Method (XFEM/GFEM) and its variants. Predicting the crack path for curved and non-planar crack surfaces is a subject of ongoing research and continuous improvement in 3D simulations. This work presents an XFEM/GFEM model for 3D crack propagation involving planar, curved, and non-planar cracks, in which a hybrid methodology for computing the crack direction is analyzed. The proposed approach for evaluating the crack direction combines the Linear Elastic Fracture Mechanics (LEFM) criterion with the Maximum Principal Stress (MPS) criterion, leveraging the strengths of each approach to address the inherent limitations of the model. The stress intensity factor (SIF) for each crack opening mode is calculated using the Displacement Correlation Method (DCM), and the proposed model can be applied to predict the crack paths of both pure and cohesive cracks in three dimensions.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13625 2025-12-18T20:45:17+00:00 Erick Battiston erickbatiston2210@gmail.com Ricardo Favaro ricardo.Favaro@sp.senai.br Hugo Sakai Idagawa hugo.Sakai@sp.senai.br

Auxetic structures have aroused significant interest in the scientific community due to their unique properties of lateral expansion when subjected to tensile forces, and transverse deformation when subjected to negative Poisson's ratio. This article offers an analysis of these structures, where the theoretical foundations of the negative Poisson's ratio behind auxetic behavior are explored, including the mechanical properties that confer this characteristic. Using SolidWorks Simulation software, mechanical simulations of re-entrant auxetic structures were carried out, and we obtained displacement results equal to 0.690, 0.910 and 1.01 mm, under loads of 150, 200 and 220 kgf respectively. These conditions were reproduced experimentally, resulting in Poisson's ratio at -0,700, -0,622 and -0,500 under each of the previously mentioned loads. Auxetic structures have an example of application in Aeronautics, where they are used in aircraft components to improve impact resistance and energy absorption, increasing passenger safety. In the area of ​​robotics, they also have application to create flexible and resistant components, such as joints and coatings, which can withstand deformations without losing structural integrity.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13641 2025-12-18T20:45:55+00:00 Gabriel Costa Rodrigues gcrmestre@gmail.com Carlos Eduardo de Souza ce.souza@ufrgs.br Bruno Klahr 00206849@ufrgs.br Daniel Milbrath de Leon daniel.leon@ufrgs.br

To execute fatigue analysis of offshore structures, it is important to adequately determine the stresses field, especially on welds. Welded regions tend to have significant stress concentration. To deal with this problem, some international standards suggest the hot spot stress (HSS) method to estimate the stresses near welds, including guidelines for finite element analysis. Two strategies of HSS calculations are considered here: surface-extrapolated stress and through-thickness linearized stress. Recommendations exist for elements sizes of mesh depending on element type, geometry and method for HSS calculation. In this work, the calculations are applied initially to a simple flat model with different meshes, with varying element sizes and proportions between the three dimensions. Next, the strategies are applied to more complex structure, representing a bell mouth used in offshore oil platforms. Being mostly a cylindrical geometry, meshing is controlled mainly by the number of elements in thickness as well as element size in circumferential and axial direction. Different loads were applied on each case to verify mesh sensitivity relation with load type. Results indicate different levels of effect on HSS results for each direction of mesh refinement. The relation of sensitivity of the mesh with the load applied is shown and discussed.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13657 2025-12-18T20:46:30+00:00 Amanda Borges Oliveira engcivil.amandab@gmail.com Paula de Oliveira Bezerra Diniz engpaula.uerj@gmail.com Gilvan Lunz Debona gilvanld@ufrrj.br José Guilherme Santos da Silvaa jgss@uerj.br

This research work aims to investigate the dynamic structural behaviour of pedestrian footbridges, when subjected to dynamic actions associated with vandalism (rhythmic human loads), based on the development of experimental tests and numerical modelling. The investigated pedestrian footbridge is associated to a simply supported concrete structure with span of 24.4 m, currently used for pedestrian crossing and located at the Faculty of Engineering (FEN) of the Rio de Janeiro State University (UERJ). Initially, an experimental modal analysis was performed on the investigated footbridge, in order to identify and assess the global dynamic behaviour of the system. After that, several loading scenarios were idealized to carry out experimental forced vibration tests on the structure, with the participation of up to 20 people, aiming to simulate dynamic excitations associated with vandalism loads. Subsequently, a finite element model was developed and calibrated based on experimental results, considering traditional mathematical functions, corresponding to only-force models, in order to represent human-induced dynamic loads related to vandalism. Based on the forced vibration experimental tests and the dynamic analysis, the footbridge dynamic structural response was investigated. The results of this study indicated that excessive vibrations and human discomfort can occur from fifteen people jumping on the analysed pedestrian footbridge.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13673 2025-12-18T20:47:00+00:00 João Fernandes joao.fernandes@ctec.ufal.br Christiano Várady christiano_varady@lccv.ufal.br Joyce Tenório joyce.tenorio@lccv.ufal.br Eduardo Toledo de Lima Junior limajunior@lccv.ufal.br João Paulo Lima Santos jpls@lccv.ufal.br Rafael Dias rafael_dias@petrobras.com.br

This paper presents a Bayesian approach for identifying and classifying soil stratigraphy from Piezocone Penetration Test (CPTu) data, explicitly capturing uncertainties in CPTu-based classifications. The probabilistic model determines the most probable number of underground soil layers of different thicknesses and types. Comparisons with the Robertson chart and a fuzzy-based method were performed using real CPTu data supplied by an oil and gas operator in the Campos Basin, Southeastern Brazil. This study integrates the Robertson soil classification chart, fuzzy-based classification, and a Bayesian probabilistic method. The Bayesian technique comprises two main steps: (i) model class selection to identify the most probable number of soil layers and (ii) system identification to estimate layer thickness and classify soil types. In the classical approaches, the layer thickness was determined using a Kernel Density Function technique. The probabilistic model merges prior knowledge with site-specific CPTu data from tests conducted in the Campos Basin. The Bayesian classification technique demonstrated strong capability for identifying soil stratigraphy, and determining the most probable layer boundaries. This approach begins with statistically significant boundaries and incrementally refines the resolution, effectively filtering out potential noise from the raw CPTu measurements. Comparative assessments show that the Bayesian model closely aligned with the outcomes from the Robertson soil classification chart and a fuzzy-based method, offering comparable accuracy while incorporating explicit uncertainty quantification. Notably, the Bayesian framework effectively adapts to varying subsurface conditions, providing a more thorough understanding of stratigraphic transitions than deterministic methods. As the number of model classes increases, the approach achieves finer stratification, confirming its utility for complex depositional environments. The probabilistic underpinnings of the method are especially beneficial in offshore petroleum applications, where reliable site characterization is critical. Overall, the results highlight the robustness and versatility of the proposed Bayesian strategy in improving geotechnical assessments and guiding informed design decisions.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13632 2025-12-18T20:45:35+00:00 Sakthivel Munikrishnan 03093921953@udesc.br Bruno Yan dos Anjos bruno.anjos@edu.udesc.br Gustavo Trindade Choaire gustavo.choaire@udesc.br Luiz A. Hegele Junior luiz.hegele@udesc.br

In the present work, we develop a two-dimensional, GPU-accelerated regularized lattice Boltzmann method (LBM) using CUDA to investigate the flow past a cylinder. The solver is based on the Bhatnagar-Gross-Krook (BGK) collision model with D2Q9 lattice stencils. A moments-based curved boundary condition is formulated to handle complex geometries on a uniform Cartesian grid, and the numerical stability is significantly enhanced through a regularization procedure. Lattice nodes in the vicinity of the physical curved boundary are identified as boundary nodes, where solid boundary conditions are applied. The accuracy and capability of the developed solver in handling curved geometries are verified using the benchmark problem of flow past a cylinder. The framework successfully captures key flow features such as flow separation, reattachment, wake formation, and unsteady vortex dynamics. Additionally, flow parameters including lift and drag coefficients, surface pressure distribution, and vortex shedding frequency (Strouhal number) are evaluated and compared against existing literature, demonstrating the accuracy of the method. The solver is fully implemented on the CUDA platform, leveraging the inherent locality of the LBM for high parallel efficiency. The GPU-accelerated implementation achieves significant reductions in computational time compared to its CPU-based counterpart.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13648 2025-12-18T20:46:12+00:00 Silvia Corbani corbani@poli.ufrj.br André Maués Brabo Pereira andre@ic.uff.br Gabriel Maurício Macena dos Santos gabrielmm@poli.ufrj.br Leonardo Oliveira Rodriguez leonardo_rodriguez@poli.ufrj.br

Teaching linear elastic fracture mechanics faces challenges related to conceptual understanding, computational modeling, and the time required. The modeling process involves predicting discontinuities, crack mapping and mesh discretization, particularly in finite elements. The objective of this work is to develop a Python code to numerically analyze cracked components, based on the hypotheses of linear-elastic fracture mechanics with axial symmetry. The two-dimensional analysis is performed using quadratic tetrahedral finite elements. The methodology includes the automatic generation of the mesh mapping the elements around the crack tip, which are usually called rosette. The user of this software can choose whether to numerically analyze a rosette of 1 to 3 rings. The first ring is composed of Quarter Point elements. The calculation of the stress intensity factor (SIF) is done using the J-Integral methods and the displacement correlation method. A user interface is developed using the Tkinter library in the Python programming environment. The Python program helps students interested in the implementation to understand the process, and the Tkinter interface is important to familiarize with the data inputs of problems involving fracture mechanics. Validation examples are compared with the literature to verify the code.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13664 2025-12-18T20:46:43+00:00 juan vallejo juanfevallejospina@gmail.com Letícia Fleck Fadel Miguel letffm@ufrgs.br Jesus Daniel Villalba Morales jesus.villalba@javeriana.edu.co

Tuned Liquid Dampers (TLDs) are passive structural vibration control devices whose efficiency and low cost make them a promising solution in civil engineering applications. Their implementation is particularly beneficial in slender or high-rise structures, in which dynamic loads—such as wind or seismic excitations—can compromise structural safety, the integrity of non-structural components, and occupant comfort. The operational principle is based on tuning the fundamental sloshing frequency of the liquid to match the natural frequency of the structure’s dominant mode, thereby generating inertial forces on the tank walls that counteract the structural motion.This study numerically evaluates the performance of a rooftop-mounted TLD in a steel Moment-Resisting Frame (MRF) using OpenSees. The analysis includes both harmonic resonant excitations and representative seismic records. The TLD is modeled using an Equivalent Mechanical Model (EMM) based on Housner’s analogy, which captures the liquid’s nonlinear behavior through a coupled mass-spring-damper system.Results indicate that the EMM provides accurate predictions of the TLD’s response under conditions of low excitation amplitude and high structural damping, achieving notable reductions in dynamic response. However, a progressive loss of efficiency is observed as the liquid’s depth-to-length ratio (h/L) increases, underscoring the importance of this parameter in TLD design.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13680 2025-12-18T20:47:21+00:00 Juliane Gonçalves juliane.goncalves@ufjf.br Ana Carolina Martins da Silva anacarolina.martins@estudante.ufjf.br

This study aims to evaluate the effect of floor plan geometry of reinforced concrete buildings on the soil-structure interaction (SSI) mechanism. Two three-dimensional finite element models were developed using the SAP2000 software: (i) a model with a square floor plan and (ii) a model with a rectangular floor plan. Both fixed (rigid) supports and elastic (spring) supports were employed in the analyses to simulate foundation behavior. The results indicate that both models exhibit similar SSI responses, with noticeable redistribution of internal forces when interaction is considered. Peripheral columns experienced increased axial demands, while central columns showed a reduction. Additionally, there was an increase in positive bending moments at mid-spans and negative moments at the peripheral supports of the central ground floor beam. These findings highlight that neglecting settlement effects - by disregarding SSI in the design process - can result in bending moment diagrams that deviate from design assumptions, potentially leading to localized plastic hinges in beams. It is also noteworthy that the rectangular floor plan exhibited higher differential settlement values, making the influence of SSI even more significant in the structural and foundation design.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13623 2025-12-18T20:45:14+00:00 Magno Mota magnomota@coc.ufrj.br Eduardo M. R. Fairbairn eduardo@coc.ufrj.br Henrique C. C. Andrade henriqueconde@coc.ufrj.br Mariane R. Rita mariane_rita@numats.coc.ufrj.br Gustavo L. X. Costa gustavo.costa@coc.ufrj.br Jean-Louis Tailhan jean-louis.tailhan@univ-eiffel.fr

This paper reports the employment of parallelization techniques in the implementation of a 3D probabilistic explicit cracking model for concrete. The mentioned model is based on the use of interface elements to explicitly represent cracks and naturally has a high computational cost, which justifies the importance of developing parallelization strategies to accelerate the analyses. Parallelization was considered in the most time-consuming tasks in the finite element code: assembly of the stiffness matrix and residual force vector, and solution of the linear equation system. The results show that significant reductions in simulation time can be obtained with the parallelized code.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13639 2025-12-18T20:45:51+00:00 Matheus de Araujo Siqueira matheusdearaujo211@gmail.com Bruno Furieri furieribruno@gmail.com Jean-Luc Harion jean-luc.harion@imt-nord-europe.fr Jane Meri Santos jmerisantos@yahoo.com.br

Diffuse emissions of granular material from storage piles open to the atmosphere represent an important source of pollution by industrial sites. Thus, to better deal with its impacts, a good understanding of the airflow around piles at small scales is vital, for example, at the characteristic scale of an industrial site. In this sense, Computational Fluid Dynamics (CFD) numerical simulations are largely applied. However, recent works have reported non-symmetric results for the airflow around the storage pile. This is not expected for this kind of geometrical case. The present work evaluates the influence, in terms of symmetry, of numerically simulating the airflow around the storage pile in a steady and unsteady state, using the open source software OpenFOAM-v2312. It was shown that the cubic building presence induces several instabilities in the airflow. Such instabilities cause convergence problems in the steady state solution and produce non-symmetric results. On the other hand, through an unsteady simulation, it was possible to capture the instabilities well and thus obtain averaged symmetric results. Second-order upwind methods were used.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13655 2025-12-18T20:46:28+00:00 Jose Fabiano Martins Assuncao assuncao.fabiano@gmail.com Leonardo Scardua lascardua@ifes.edu.br

The selection of rolling sequence types in the Hot Strip Mill (HSM) in ArcelorMittal Tubarão (Largest steel producer in Latin America and global market leader) is traditionally based on empirical evaluation by operational experts. However, the complexity arising from order variability, slab yard dynamics, and operational constraints limits the efficiency of manual decision-making. This study proposes a methodology to automate the classification of sequence types, based on feature analysis of approximately 360 empirically assessed sequences. Preprocessing and feature engineering techniques were employed to support supervised learning models trained to categorize four main sequence types. The results show consistent performance, with high levels of accuracy, sensitivity, and specificity, highlighting the potential of the approach to improve in 10% process efficiency and compliance with key production KPIs such as decarbonization and customer satisfaction.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13671 2025-12-18T20:46:57+00:00 Lucas Cruz lucas.senna.lsc@hotmail.com Elisabeth Junges Laure elisabeth.junges@ufes.br Lorenzo Augusto Ruschi e Luchi lorenzo.luchi@ufes.br

Verifying structural deformations is essential to ensure the performance and safety of the designed system. For prestressed concrete beams, deflection calculations involve additional particularities compared to reinforced concrete beams. These include, for instance, the need to account for the effect of prestressing force on the concrete and the presence of prestress-induced hyperstatic effects in continuous beams. In cases of partial prestressing, the calculation of the immediate deflection component requires consideration of the stiffness reduction due to concrete cracking, just as with reinforced concrete beams. Although prestressed beams typically exhibit lower levels of cracking compared to reinforced concrete beams, it is still necessary to account for this effect in deflection calculations. Therefore, it is essential to comprehend how various design standards address these elements. This study aims to compare simplified methods for calculating short-term deflections in simply supported and continuous prestressed concrete beams with partial prestressing levels, following the provisions of the Brazilian standard NBR 6118 [1] and the fib Model Code [2]. Deflection values are computed using both normative models for a set of simply supported and continuous prestressed concrete beams designed with partial prestressing. A comparative analysis is presented, discussing the differences observed between the methods.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13630 2025-12-18T20:45:26+00:00 Adriano Cortes adriano@nacad.ufrj.br Roberto M. Velho robertovelho@cos.ufrj.br Rodrigo de S. Luna rluna@cos.ufrj.br Thiago H. N. Coelho tcoelho@cos.ufrj.br Renato N. Elias rnelias@nacad.ufrj.br Alexandre Evsukoff alexandre.evsukoff@coc.ufrj.br Herve Gross herve.gross@totalenergies.com Mauricio Araya mauricio.araya@totalenergies.com Fernando A. Rochinha faro@mecanica.coppe.ufrj.br Alvaro L. G. A. Coutinho alvaro@nacad.ufrj.br

Carbon Capture and Storage (CCS) is essential for mitigating climate change by permanently storing industrial CO2 emissions in subsurface formations [1]. As net-zero targets gain urgency, accurately predicting injected CO2 behavior over long timeframes becomes critical for ensuring storage integrity and regulatory compliance. Simulating CO2 plume migration presents significant computational challenges. Conventional numerical simulators, while accurate, require prohibitive computational resources when modeling the centuries-long timescales relevant to CO2 storage, especially for uncertainty quantification or optimization studies requiring multiple simulations. Deep learning surrogate models offer orders-of-magnitude speedup while maintaining reasonable accuracy [2].Current literature features two main surrogate modeling approaches: Fourier Neural Operators (FNO) [2] and Graph Neural Networks (GNN) [3]. FNO models perform well on structured domains but are limited to Cartesian grids, inadequately representing complex geological features. In contrast, GNN approaches like MeshGraphNet (MGN) [4] can naturally handle unstructured meshes, better representing heterogeneous subsurface environments and geological boundaries critical to CO2 migration.We present our MGN-LSTM implementation for temporal prediction of CO2 plume migration in faulted reservoirs [3], incorporating algorithmic improvements for enhanced prediction stability and efficiency. Our key contribution is integrating physics-aware loss terms from FNO literature [5] into the GNN framework. By incorporating mass conservation, phase behavior, and Darcy flow principles directly into the loss function, we achieve physically consistent predictions even with limited training data. These physics-aware constraints improve model generalization to unseen geological configurations.References[1] IPCC. Special Report on Carbon Dioxide Capture and Storage. CUP, 2005.[2] Wen, G. et al. Towards a predictor for CO2 plume migration using deep neural networks. Int. J. Greenhouse Gas Control, 2021.[3] Ju, X et al Learning CO2 plume migration in faulted reservoirs with Graph Neural Networks. Computers & Geosciences, 193, 2024.[4] Pfaff, T. et al. Learning mesh-based simulation with graph networks. ICML, 2021.[5] Badawi, D., Gildin, E. Neural operator-based proxy for reservoir simulations considering varying well settings, locations, and permeability fields. J. Petroleum Science and Engineering, 2023.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13646 2025-12-18T20:46:08+00:00 Jackson da Silva Rocha Segundo jacksonsrsegundo@gmail.com Ricardo Azoubel da Mota Silveira ricardo@ufop.edu.br Rafael Cesário Barros rafaelcesario@hotmail.com Dalilah Pires dalilah@ufsj.edu.br Ígor José Mendes Lemes igor.lemes@ufla.br Caroline Aparecida Ferreira caroline.af@aluno.ufop.edu.br

Wooden structures are increasingly used in civil construction. Due to its ecological and renewable nature, and its wide variety of byproducts, this material has stood out as an intelligent solution for structural elements. Therefore, knowing its behavior is of utmost importance to ensure safety for designers. Due to the combustibility of this material, understanding the consequences of exposure to a fire situation is essential to ensure the safety of the building. Therefore, this research seeks to evaluate the thermostructural behavior of a wooden beam in a fire situation via CS-ASA/FSA. In comparison with the literature, the ability of the computational tool to perform this type of analysis is evident. The results obtained identify the importance of temperature advances in the stiffness and resistance capacity of the wooden beam. Additionally, the importance of understanding the behavior of wood in these scenarios is evident.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13662 2025-12-18T20:46:40+00:00 Lucas Araújo Rodrigues da Silva araujolucasrs@usp.br André Teófilo Beck atbeck@sc.usp.br Luis Gustavo Lopes Costa luis.lopes.costa@usp.br Jochen Köhler jochen.kohler@ntnu.no

In this paper, we explore the interdependence between quality control, represented by inspection frequency, and optimal redundancy in structural systems. A risk-based optimization of a Daniels system is performed, considering inspection intervals throughout the design life. System reliability is evaluated using a model where loads are represented as pulses of random intensity, and failures due to non-technical factors are modeled as pulses governed by a latent hazard rate λL . Both brittle and ductile material behaviors are considered, along with varying correlation coefficients between element strengths. The results reveal a clear dependency between the optimal number of inspections (ninsp ) and the optimal number of elements (nc ). As λL increases, both ninsp and nc increase accordingly. Moreover, the results are strongly influenced by the strength correlation between elements and the material behavior. The findings highlight the role of latent failure probability as a conceptual link between quality control and structural reliability.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13678 2025-12-18T20:47:17+00:00 Andrey Alexandre de Souza Basilio andreyalexandre16@gmail.com Daniel Cruz Cavalieri daniel.cavalieri@ifes.edu.br

This paper proposes a multi-agent system based on large-scale language models (LLMs) for automated skin cancer diagnosis from images. The architecture includes segmentation, classification, and report generation steps, integrating specialized agents based on models such as Gemini-2.5-pro and GPT-4o. The approach was evaluated using bootstrap sampling with five replicates of size 80, highlighting the efficiency of the visual agents and the consistency of the diagnoses generated. Preliminary results indicate the potential of integrating computer vision and natural language to support medical diagnosis.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13621 2025-12-18T20:45:09+00:00 SAULO Silvestre DE CASTRO saullo9@yahoo.com.br Hugo M. Leão hugomleao@yahoo.com.br Adalberto Horácio de Oliveira Mendes adalbertohom@gmail.com Roque Luiz da Silva Pitangueira roque@dees.ufmg.br

Machine learning (ML) models have emerged as promising alternatives to classical constitutive formulations for representing quasi brittle materials, such as concrete, rock, and composite media. Unlike traditional analytical approaches, ML models are trained directly on experimental or simulated data, enabling them to capture complex nonlinear behaviors—such as stiffness degradation, progressive cracking, and brittle-to-ductile transitions—in a more adaptive and generalizable manner. Their inherent flexibility allows for the incorporation of material variability and the customization of constitutive responses according to boundary conditions and loading history.However, the integration of ML-based constitutive models into finite element simulation environments demands substantial modifications to the underlying software architecture. Since such models operate through predictions made by neural networks or similar algorithms, they require dedicated structures for storing and updating internal variables, as well as efficient communication mechanisms with the numerical core to ensure the stability and performance of the analysis.This work presents the strategies adopted to adapt the INSANE system—originally developed in Java—to support ML-based constitutive modeling. Interoperability mechanisms with Python environments were implemented, alongside robust runtime data exchange interfaces and modifications to the constitutive integration logic, while preserving the modularity of the system. The adopted approach enables future extensions to various ML architectures and enhances INSANE’s predictive capabilities in simulations involving brittle and heterogeneous materials.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13637 2025-12-18T20:45:48+00:00 Gustavo Kenzo Tamura kenzotamura@usp.br André Teófilo Beck atbeck@sp.usp.br

In recent years, extreme climate events such as floods, wildfires, heatwaves, and prolonged droughts have been observed with increasing frequency, intensity, and geographical spread. Due to this, there has been growing interest in sustainability, particularly in structures, since the construction and operation of buildings account for approximately 37% of global CO₂ emissions. Thus, dematerialization is an extremely relevant topic for the conscious and rational use of natural resources in buildings. However, current Brazilian technical standards do not reflect this concern, being potentially conservative without significant gains in structural safety. To align these two interests, reliability theory can be applied, ensuring structural safety while reducing the environmental impact of constructions. The initial goal of the reliability-based calibration of partial safety factors was to achieve more uniform safety levels across different structural designs. As an additional benefit, it was possible to observe a reduction in design loads and, consequently, in construction material consumption. In this study, CO₂ emission reductions in slabs were evaluated by comparing a design using standard safety factors from design codes and using reliability-based calibrated factors. The Brazilian tool Sidac was employed to estimate CO₂ emissions. Using the calibrated partial factors, an average 7.65% reduction in CO₂ emissions and a 7.41% cost reduction were achieved in the design of case-study slabs.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13653 2025-12-18T20:46:23+00:00 RAFHAELA LINHARES rafhaela.linhares@alunos.ufersa.edu.br Raimundo Gomes de Amorim Neto raimundoamorim@ufersa.edu.br Vinícius Navarro Varela Tinoco navarro@ufersa.edu.br Flaviana Moreira de Souza Amorim joflaviana@yahoo.com.br

Present in the curriculum of the Civil Engineering program, Structural Mechanics is one of the most relevant and complex subjects in the degree course, given the breadth of its concepts and methods. Structural Analysis is the branch of mechanics that studies structures, determining the internal forces and deformations they undergo when subjected to external loads. Structural elements are the components that make up a structure; one of them is the beam, which is a slender structural element that supports forces acting perpendicular to its longitudinal axis.Beams are classified according to how they are supported: a simply supported beam has support at both ends, with one fixed and the other movable (or roller); a cantilever beam has a fixed end (encastre) at one side and the other end free; an overhanging beam has one or both ends extending beyond its supports (Hibbeler, 2010).Therefore, this project aims to develop a mobile application for the Android operating system to assist Civil Engineering students in the structural analysis of isostatic beams during the Structural Mechanics course. Using the visual, intuitive, block-based programming environment of MIT App Inventor, the application BeamCalc was created.The software calculates support reactions and displays shear force and bending moment diagrams, as well as the maximum moment of isostatic beams under various types of loads. Throughout this work, important concepts were analyzed, such as types of beam supports, their respective support reactions, and the equilibrium equations. To validate the results, a comparison was made with those obtained using the Ftool software, showing the application to be effective.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13669 2025-12-18T20:46:54+00:00 Anderson Gomes Girardi andersongomes@ufg.br Andreia Aoyagui Nascimento aanascimento@ufg.br Felipe Pamplona Mariano fpmariano@ufg.br

To enhance understanding of the conversion of wind kinetic energy into mechanical work, this study presents the design and construction of vertical-axis Savonius wind turbine models for educational use in a wind tunnel. The work emphasizes the use of low-cost materials and a simplified manufacturing process. Two turbine models with distinct configurations were developed. The first model was constructed using PVC pipes, featuring overlapping blades, a single stage, and a defined central shaft. The second model was fabricated from acrylic, with non-overlapping blades, a two-stage configuration, and no apparent central shaft; the blades are mounted on acrylic disks, which in turn hold shaft tips that connect to either a torque measurement device or an electric generator. The two-stage design of the second model enables automatic startup of rotation, and the absence of a central shaft reduces the overall weight of the turbine, although it poses challenges for balancing. Rotational speed measurements were performed using a tachometer, while torque was measured using a custom-built static torque meter. The results include the tip speed ratio relative to wind speed and the torque coefficient, i.e., the efficiency curves of each wind turbine model. These are compared with results reported in the literature.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13628 2025-12-18T20:45:23+00:00 T. Doca doca@unb.br R.M. Silva Filho filho.ricardo@aluno.unb.br

This study evaluates the capability of the Arruda–Boyce elasto-viscoplastic model to predict the monotonic compressive response of the PC/ABS 80:20 polymer blend. A finite element model of a standard compression specimen, following ASTM D695, was developed in the commercial package ABAQUS. The model incorporates geometric, contact, and material nonlinearities via a custom VUMAT routine. It effectively captured the material’s nonlinear response, including strain softening, and showed good agreement with experimental stress–strain data. The work emphasizes the superiority of force–displacement data for constitutive calibration and demonstrates the potential of population-based optimization methods using a one-dimensional viscoplastic model implementation. These findings establish a solid basis for further extension of the model to more complex loading scenarios.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13644 2025-12-18T20:46:02+00:00 Igor Braz igor.gonzaga@coc.ufrj.br Michèle Schubert Pfeil mpfeil@coc.ufrj.br Wendell Diniz Varela wendell@coc.ufrj.br

Human–Structure Interaction (HSI) refers to the dynamic interplay between pedestrians and footbridge structures, forming a coupled mechanical system that may be represented by a single degree of freedom model (SDoFM) with modal parameters that differ from those of the empty footbridge. This study aims to present fitted equations to estimate the modal parameters of the coupled system, derived from numerical simulations emulating free vibration tests of occupied footbridges. Structural responses are evaluated using three methodologies, under pedestrian induced walking loads: (i) solving the coupled multi-degree-of-freedom model (MDoFM) of the footbridge–pedestrian system; (ii) solving the SDoFM of the coupled system, using the proposed equations; and (iii) solving the SDoFM of the empty footbridge, disregarding HSI effects. Results show neglecting HSI leads to significantly higher vibration amplitudes than those displayed by the coupled systems analyses, while MDoFM and SDoFM yield similar responses, demonstrating the practical relevance of the proposed HSI-SDoFM approach.

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13660 2025-12-18T20:46:36+00:00 GERSON H. DOS SANTOS gsantos@utfpr.edu.br Anderson Lemes da Silva anderson.2004@alunos.utfpr.edu.br Thiago Antonini Alves antonini@utfpr.edu.br

Earth-Air Heat Exchangers (EAHE), also known as Ground-to-Air Heat Exchangers (GAHE), are sustainable and cost-effective systems designed to improve building energy efficiency by utilizing the thermal characteristics of the soil. The core principle behind EAHE is that the soil beneath the surface maintains a relatively constant temperature throughout the year, depending on the local climate and depth. This stability makes it an ideal medium for passive heat exchange, reducing the reliance on conventional heating and cooling systems. The system consists of underground pipes, arranged in horizontal or vertical configurations. During operation, air is drawn through these buried pipes, where it either gains heat from the warmer soil during winter or releases heat into the cooler soil during summer, effectively tempering the airflow before it enters the building. This process significantly reduces the energy needed for indoor climate control, leading to lower utility bills and a smaller environmental footprint. However, the performance of an EAHE depends heavily on the soil's hygrothermal properties, which influence heat and moisture transfer within the ground. Furthermore, in cold climates, the exchanger's performance can be increased by exposing part of the piping to solar radiation through a bypass. Furthermore, in cold climates, the exchanger's performance can be increased by exposing part of the piping to solar radiation through a bypass. To address this, a 2D model was developed to analyze the exchanger's performance. A modification in the tube circuit is proposed to take advantage of the effects of solar radiation on air heating. Experimental validation was performed through a prototype built at the Federal University of Technology of Paraná (UTFPR).

2025-12-01T00:00:00+00:00 Copyright (c) https://publicacoes.softaliza.com.br/cilamce2025/article/view/13676 2025-12-18T20:47:11+00:00 Victoria Reis victoria.dias.reiss@gmail.com Felipe Ramos de Oliveira felipe.oliveira@coc.ufrj.br Nelson Francisco Favilla Ebecken nelson@ntt.ufrj.br

This study explores the use of a supervised Retrieval-Augmented Generation (RAG) pipeline for automatic misinformation classification in Brazilian Portuguese. The method combines semantic retrieval based on dense embeddings with traditional machine learning classification using TF–IDF and Support Vector Machine. Experiments were conducted on the largest publicly available Brazilian Portuguese fake news dataset, previously consolidated by the authors from multiple sources, and performance was compared with results from the literature using classical methods such as SVM. The results indicate that while supervised RAG provides competitive performance, its gains over traditional approaches may be limited when the dataset is balanced and linguistically homogeneous. A detailed error analysis is presented, and the potential of supervised RAG in more challenging and low-resource scenarios is discussed.

2025-12-01T00:00:00+00:00 Copyright (c)