Predicting defect-driven strain localization in additively manufactured materials remains challenging because experimental datasets are often limited, expensive to generate, and characterized by complex defect topologies. Conventional machine learning models can provide accurate predictions but often lack physical interpretability and uncertainty quantification. To address these limitations, this study proposes a physics-guided and explainable surrogate learning framework for topology-aware damage evolution and localized strain prediction. The framework integrates physics-guided defect descriptors, surrogate-assisted data augmentation, Gaussian Process Regression (GPR), and SHapley Additive exPlanations-based explainable artificial intelligence (AI) to enable accurate, interpretable, and uncertainty-aware predictions under small-data conditions. Physically meaningful descriptors, including topology damage, defect concentration, inter-defect distance, heterogeneity, and damage entropy, were formulated to capture nonlinear defect interactions governing localized deformation. The proposed GPR model achieved an R² of 0.895, with a mean absolute error (MAE) of 0.017 and root mean square error (RMSE) of 0.020, outperforming the benchmark XGBoost model (R² = 0.794, MAE = 0.031, RMSE = 0.041). Explainability analysis identified heterogeneity and topology damage as the dominant factors influencing strain localization. The proposed framework provides a computationally efficient, physically interpretable, and uncertainty-aware approach for defect-sensitive strain prediction, supporting reliable small-data modelling and future digital twin applications in additive manufacturing.
Internal defects, such as pores and inclusion-like features, may compromise the structural quality and reliability of polymer piezocomposites. This study appraises the influence of repeated extrusion on the internal defect characteristics of barium titanate (BTO) and reduced graphene oxide (rGO) polymer piezocomposites using X-ray computed tomography (CT). Six specimens were examined: one BTO and one rGO specimen produced after each of one, two, and three extrusion passes. Reconstructed three-dimensional CT datasets were analysed in VGStudio MAX to quantify pore populations in the BTO specimens and inclusion-like features in the rGO specimens, together with their size- and morphology-related distributions. The BTO pore count decreased from 980 after one extrusion pass to 74 after two passes and 8 after three passes, while the number of detected inclusion-like features in the rGO specimens decreased from 1819 to 314 and 307, respectively. These results show that repeated extrusion was associated with reduced detectable defect populations and changes in defect distributions. The main scientific contribution of this work is the quantitative three-dimensional comparison of processing-induced microstructural changes in two polymer piezocomposite systems using CT-derived defect descriptors. As only one specimen was analysed per condition, further measurements are needed to assess repeatability and statistical significance.
This paper provides an analysis of the transverse sensitivity of a strain gage made with cold rolled, Cu-Ni foil (constantan) by examining the individual contributions of the grid lines and end loops of the strain gage. The analysis utilizes two types of finite element models. Two dimensional models of the electrical current flow through the grid lines and end loops are used to determine the electrical resistance of each component. Three dimensional models of the strain gage attached to a rigid substrate subjected to transverse loading are used to determine the surface strain in each component. Results of both models are combined to determine strain gage transverse sensitivity - replicating standards for physical measurement of transverse sensitivity. Several end loop configurations are considered including geometry (rectangular, curved, and H-shape); size (length and width), and foil material property (modulus of elasticity and electrical resistivity). The factors contributing most to transverse sensitivity of the strain gage are the electrical resistance and surface strain in the transverse direction of the end loops. Good correlation is found between analytical model results and physical test results.
Weld joint failures remain a significant concern in the manufacturing and construction sectors, particularly in developing countries where artisanal welding dominates. This study assessed the knowledge of welding and fabrication artisans and professional welders in Ghana regarding the causes of weld joint failure. A structured questionnaire was administered to 379 respondents, capturing socio-demographic characteristics, technical knowledge across design, material selection, fabrication, and service-related domains, as well as perceived training needs and causes of weld failures. Descriptive, inferential, and regression analyses were applied to examine knowledge levels and their predictors. Findings revealed that welders demonstrated high knowledge in design, material, and fabrication domains but only moderate awareness of service-related factors. Professional welders and those with formal training exhibited significantly higher levels of knowledge than artisanal welders; however, on-the-job experience positively influenced knowledge, albeit to a lesser extent than structured training. Respondents identified poor welding techniques and a combination of factors as the primary cause of weld joint failures, and strongly endorsed the need for additional training and certification programmes. Regression analysis indicated that education, experience, formal training, and adherence to welding procedures significantly predicted overall knowledge. The study concluded that formal training and procedural compliance are critical to reducing weld failures, emphasising the need for targeted interventions, certification programmes, and continuous professional development to enhance weld quality and reliability.
Formability of steel sheet metals depends primarily on mechanical properties of material that influence the material's ability to undergo plastic deformation without risk of fracturing. This article presents the use of multilayer artificial neural networks (ANNs) to evaluate the effect of uniaxial tensile test parameters on selected mechanical properties of 0.7-mm-thick DX56D+Z100-M-C-O steel sheets. Samples were cut from different areas of the coil and at different angles (0°, 45° and 90°) relative to the sheet rolling direction. Samples also came from different batches of materials. Yield strength, ultimate tensile strength and uniform elongation were determined under variable crosshead speeds between 2 and 600 mm/min. The quality parameters of the ANNs were the value of the coefficient of determination (the higher the better) and the prediction error (the smaller the better) determined for the test set. The predictive quality of ANNs was characterized by the coefficient of determination R2 = 0.7849-0.9864, depending on the ANN architecture, the output variable and type of neuron activation function used. As crosshead speed increased, elongation of samples decreased. However, as crosshead speed decreased, yield strength decreased. The influence of sample orientation on the analyzed mechanical parameters of the sheet was smaller than the effect of crosshead speed. This was also confirmed by sensitivity analysis of the input parameters.
Knowledge of the linear expansion coefficient of steel is crucial for ensuring the safety, durability, and proper functioning of welded structures. The aim of this study is to investigate the coefficient of linear expansion of S355J2 structural steel and C45 carbon steel, welded using various techniques: metal inert gas welding, tungsten inert gas welding and manual metal arc welding. Linear expansion tests of steel samples were conducted using a dilatometer. The results allowed for an assessment of the effect of the welding technique and the chemical composition (carbon content) of the steel on the thermal properties of the test materials. The reference samples (materials in as-received state) achieved the highest level of thermal expansion. All welding techniques reduced the temperature-induced elongation, with the magnitude of this effect dependent on the welding technique used. The results have practical implications for the design of structural components exposed to variable temperature conditions. Understanding the effect of individual welding techniques on the properties of S355J2 and C45 steels can support the selection of material welding technologies where dimensional stability is an important quality parameter.
Fiber-reinforced polymer composite materials have gained extensive application in aerospace, automotive, marine, and civil infrastructure owing to their exceptional specific strength, stiffness, and design flexibility. However, delamination - a critical interlaminar failure mode compromises structural integrity and dynamic performance. This comprehensive study investigates the vibration behavior of carbon fiber-reinforced polymer (CFRP) composite plates subjected to varying delamination extents, laminate stacking sequences, and boundary constraints through integrated analytical and finite element methodologies. The governing differential equations are derived using the Rayleigh-Ritz energy method based on classical laminated plate theory, and numerical simulations are performed using ANSYS finite element software. The investigation examines delamination sizes ranging from 0% to 56.25% of plate area, three distinct stacking configurations ([0/90/45/90], [0/45], [0/90]), and all sides clamped (CCCC), simply supported (SSSS), cantilever (CFFF), and free edges (FFFF) boundary conditions. Results demonstrate that natural frequencies decrease systematically with increasing delamination size, with maximum reduction of 5-8% occurring for the largest delamination extent (56.25%) across all boundary condition.. Furthermore, CNT integration enhances both natural frequencies (up to 29.8% increase at 2.5 wt% CNT loading) and damping characteristics (42.1% improvement). These findings support improved design and vibration control of advanced composite structures.
Atomic force microscopy is used for characterization of biological samples, individual molecules, molecular interaction and their roles in friction and adhesion at molecular and cellular levels. The current paper aims to summarize the atomic force microscopy, its working principle, the usage of it in characterization of bone and latest developments appeared. The proposed workflow for the AFM is developed based on current observations based on articles studied in current paper and their results, their limitations and remained gaps to be solved, as well as there are also additional steps shown for educational purposes. Potential bias in elastic modulus and boundary condition errors, variability of values and lack of standardization plus possible bias sources, biases in algorithm are the limitations and problems needed to be still solved. Therefore, a combined effort of specialists in biomechanics, tissue engineering and biomaterials is necessary to be conducted further.
AA7075 thin plates are extensively used in the marine industry, particularly for the manufacturing of hydrofoil skin panels. Surface grinding is a critical finishing process for these plates, yet the optimal grinding parameters that minimize corrosion current density (Icorr) and maximize polarization resistance (Rp) are not well established. This study was conducted to determine optimal grinding settings for controlling Icorr and Rp in 3.5 wt.% NaCl solution (simulated seawater). AA7075 thin plates were ground following a design of experiments (DoE) schedule, and Icorr and Rp were measured using a CorrTest electrochemical workstation. Results showed that Icorr increased markedly with higher table speed (50 spm), feed rate (5.0 mm/min), and grinding depth (1.0 mm), while Rp decreased under the same conditions. Standardized effects analysis identified feed rate and grinding depth as the most influential factors, each with an effect of 10.94, whereas table speed had a moderate effect, and interaction terms played secondary but significant roles. Regression models demonstrated strong predictive capability, with R² and predicted R² values of 99.28% and 97.12% for Icorr, and 98.43% and 93.71% for Rp. The optimal settings were found at low table speed (2 spm), low feed rate (1.0 mm/min), and low grinding depth (0.2 mm).
Three dimensional printing technology has widely been utilized to construct a variety of complex biomimetic structures. A target structure is designed using 3D-CAD and then the CAD data is sent to the controlling unit of a 3D-printing machine to fabricate a corresponding real structure. Recently, FEA installed into 3D-CAD can be used to assess their structural integrity of the designed structures. However, it is still difficult to analyze the nonlinear mechanical responses using FEA. The aim of the present study is to develop a nonlinear FEA method to characterize the elastic-plastic deformation behaviors along with micro-damage formations of 3D-printed polymeric porous structures. It was found that the proposed FEA method can reasonably be used to predict the nonlinear behaviors of two different types of 3D-printed porous structures under compressive loading. Mechanical properties such as stiffness, fracture energy and strength were also well predicted by FEA. The micro-fracture processes of the real structures were well characterized by the damage models with FEA including the tensile cracking with the maximum principal stress criterion and the compressive crushing with the minimum principal strain criterion.
Bioinspired materials are among the most durable materials known to man. Mimicking solutions and structures observed in nature is a modern approach to modeling materials in line with sustainable development. Designers of mechanical structures are continually seeking new applications and materials that replicate natural effects. This article presents the primary natural sources of bioinspiration in the production of advanced composite materials. The focus is on discussing current advances in the production of impact-resistant composite materials. The main sources of bioinspiration for impact-resistant materials are pearl structures, insect exoskeletons, and fruit shells. Insect cuticles offer a sustainable alternative due to their exceptional stiffness, unique properties, and mechanical parameters. The use of biocomposites in the production of mechanical structures is expected to grow in the coming years due to the continuous development of new composite technologies.
This investigation conducts a comparative study on the thermo-hydraulic performance of magnesium oxide nanoparticles dispersed in transformer oil (MgO/TrO) and a hybrid nanofluid containing both magnesium oxide and copper oxide nanoparticles (MgO-CuO/TrO). Key performance metrics including Overall Heat Transfer Coefficient (UHTC), Convective Heat Transfer Coefficient (CHTC), Nusselt number (Nu), friction factor (FF), and pumping power (PP) were evaluated across relevant ranges of temperature (30-70 ̊ C) and Reynolds number. Results indicate that the hybrid (MgO-CuO/TrO) nanofluid consistently demonstrates significantly enhanced heat transfer characteristics. Compared to the MgO/TO fluid, the hybrid formulation exhibited superior UHTC up to 55% enhancement, CHTC 14-24% enhancement, and Nusselt number 4-27% enhancement. The hybrid nanofluid showed higher friction factors in the range of 8-11% and consequently required 2.2-3.8% greater pumping power under similar operating conditions. While the incorporation of CuO nanoparticles improves the thermal performance of the MgO-based transformer oil with an increased oil pumping power effect.
The efficiency of photovoltaic (PV) systems decreases as module temperature rises under high solar irradiance, leading to reduced power output and accelerated material degradation. In this study, a physics-informed neural network (PINN)–based predictive framework is developed to model the performance of passively cooled PV panels using previously published outdoor experimental data as a validated physical reference. Experimental datasets corresponding to three passive cooling configurations phase change material (PCM), aluminium fins, and a hybrid PCM–fin system reported earlier under identical operating conditions are employed as benchmark inputs for model training and validation. The proposed PINN explicitly incorporates a thermodynamically consistent temperature–efficiency relationship into the learning process, enabling physically constrained prediction of PV power output as a function of irradiance and temperature. The trained models demonstrate high predictive accuracy across all cooling configurations, with test-set coefficients of determination of approximately 0.99, 0.97, and 0.98 for the PCM, fin, and hybrid systems, respectively. When compared with a conventional artificial neural network trained under identical conditions, the PINN reduces the root-mean-square prediction error by approximately 12–18% and exhibits improved stability under previously unseen operating conditions. Overall, the results show that physics-informed learning provides a reliable and interpretable approach for modelling photovoltaic performance under passive cooling conditions. By leveraging validated experimental benchmarks rather than introducing new measurements, the proposed framework enables effective use of limited data and offers a scalable tool for comparative performance assessment and design exploration of passive PV cooling strategies in high-temperature environments.
AA7075 aluminium alloy thin plates are widely used in marine applications, particularly as skin panels for hydrofoil crafts, where surface grinding is an essential finishing operation. However, the influence of grinding parameters on mechanical behavior across different experimental run orders is not fully understood. This study examines the effects of surface grinding parameters on the mechanical properties of AA7075 thin plates, focusing on tensile strength, yield strength, elongation, and elastic modulus. Specimens were ground according to the experimental design schedule and tested using a DI-CP/V2 servo-hydraulic testing machine. Tensile and yield strengths exhibited similar fluctuation trends, with tensile strength consistently higher than yield strength. During runs 1–20, strength values were relatively low and stable, followed by a sharp increase between runs 23 and 27, where tensile and yield strengths reached approximately 170 MPa and 155 MPa, respectively. A slight reduction occurred during the mid-runs, while further peaks after run 40 indicated improved surface integrity under specific grinding conditions. Ductility analysis showed total elongation of 6–13%, exceeding uniform elongation of 4–6%, indicating a stronger influence on post-necking deformation. The elastic modulus varied between 10 and 33 GPa, showing mid-run fluctuation and stabilization toward the final runs.
Nanoindentation is a method for mapping the mechanical properties of heterogeneous materials. This paper aims to provide a review of this method, the challenges that still remain in spite of recent innovations and future recommendations. Various types of indenters and sample preparation methods, together with displacement-load curves ensure the correctness of operation process and correct manipulation of nanoindentation device. The novelty of this work lies in the proposed workflow in schematic way, that integrates Atomic Force Microscopy (AFM), Density Functional Theory (DFT) and Finite Element Method (FEM) into a single sequence. This approach enables the cross-validation between computational, experimental and quantum mechanical methods, providing a comprehensive characterization of material from macroscale to nanoscale. This workflow serves also as a standardized guideline, that may help to enhance reproducibility and interpretability for bone mechanical properties, thus it can be used for interdisciplinary projects and become a routine tool in material science and biomedical applications.
Single-point incremental forming (SPIF) is a method of forming sheet metal components in a variety of industries. SPIF involves the gradual deformation of the sheet metal using a pin tool. In this article, SPIF was used to form a Zn-Cu-Ti alloy square pyramid drawpieces with a wall angle of 60°. Zn-Cu-Ti alloy sheets are characterised by strong anisotropy associated with the hexagonal close-packed structure. The aim of the study was to determine the effect of SPIF process parameters on the strength properties of the drawpieces. Analysis of variance was used to statistically analyse the effect of SPIF process parameters on the yield strength, ultimate tensile strength and elongation of workpiece material after forming. Based on the analysis of variance, it was found that statistically significant parameters influencing SPIF-induced properties of drawpiece material (yield strength, ultimate tensile strength and elongation) were workpiece orientation, orientation of samples taken for testing in relation to the sheet rolling direction and tool rotational speed. Step size significantly affects the yield strength and ultimate tensile strength of drawpiece material.
In this paper, the effect of sheet pre-deformation on the change of the surface roughness parameters and friction coefficient value is investigated. For this purpose, strips of AISI 430 ferritic stainless steel with deep drawing quality (DDQ), measuring 0.8 × 25 × 500 mm, were pre-deformed using a uniaxial tensile test for five different true strain values. The correlation between the surface roughness parameters and hardness with the frictional conditions of the tested strips was investigated in the bending under tension test. The results revealed that the friction coefficient determined for all pre-deformed strips increased as the level of true strain also increased. An increase in the plastic deformation of sheets under the uniaxial tensile stress state causes a nearly linear increase in the value of basic amplitude parameters of surface roughness, however, the hardness tended to present a constant increase for deformations close to uniform elongation. Furthermore, scratches and severe wear occurred on the surface of the strips and intensified with increasing roughness.
In this paper, thermal performance of transformer oil enhanced with addition of multi-walled carbon nanotubes (MWCNTs) in a plate-fin heat exchanger is experimentally determined. In this investigation, a custom-built test rig, is employed to determine the overall heat transfer coefficient (UHTC) and friction factor. Thermo-physical properties of MWCNT based transformer oil are experimentally determined in the laboratory. From this investigation, the key findings indicate that adding MWCNTs up to 0.008% concentration significantly boosts thermal performance of oil, achieving notable increases in UHTC up to 34% as a function of mean bulk temperature of the nano-fluid taking mass flow rate a parameter. Likewise, there is 20% enhancement in convective heat transfer coefficient (CHTC) as a function of mass flow rate particularly at higher flow rates. However, this enhancement is coupled with few drawbacks in the context of fluid dynamics, such as the nanofluids exhibiting 3.8% increase in friction factor at low Reynolds numbers leading to 4.5% extra pumping power requirement when compared to the base fluid. MWCNT-transformer oil based nanofluids offer improved heat dissipation capacity and highlights the potential of these nanofluids to significantly improve thermal system performance, particularly at high flow rates, without incurring excessive pumping power demands.
Friction studies in sheet metal forming processes are essential for developing appropriate forming technology. This paper presents the results of experimental studies of friction occurring at the drawbead. A 0.8-mm-thick low-carbon DC04 steel sheet was used as the test material. Friction studies were conducted under machine oil lubrication conditions. Due to the anisotropy of test material properties, the friction tests considered strip specimens cut longitudinally and transversely to the sheet rolling direction (RD). The obtained results allowed the determination of the effect of friction process parameters on the coefficient of friction (CoF). For both sample orientations, increasing the drawbead height led to a decrease in the CoF. Samples cut transversely to the RD showed higher CoFs compared to specimens cut in the RD. Based on scanning electron microscopy micrographs, it was determined that the primary friction mechanisms were flattening and microploughing of sheet metal surface.
Titanium dioxide (TiO2) is extensively employed due to its distinctive attractive thermal and physical characteristics. Various techniques, encompassing both single-step and two-step methods, have been utilized by researchers for the preparation of TiO2-based nanofluids. In its natural state, TiO2 is found in three crystalline forms: anatase, brookite, and rutile. However, the direct application of nanoparticles in heat transfer scenarios presents a considerable challenge, compelling scientists to seek out stable nanofluid preparation techniques. Nanofluids have gained recognition as promising thermodynamic fluids, largely because of their impressive attributes in thermal convection, conduction, stability, and heat transfer. An exponential surge in research has been observed concerning their thermo-physical properties, potential advantages, and applications. While numerous reviews strive to deliver comprehensive summaries on the preparation, characteristics, heat transfer, and application performance of diverse nanofluids, the sheer volume of existing literature makes this a daunting task. Therefore, a selective yet thorough summary that focuses on a specific aspect of a particular nanofluid is highly valuable. This review concentrates on the heat transfer characteristics of TiO2-based nanofluid, which is regarded as one of the most practical options for real-world applications owing to its superior dispersibility, chemical stability, and non-toxic properties. Ultimately, this review paper aims to provide a wide-ranging overview of the research progress in the heat transfer applications of TiO2-based nanofluids.
