Research Center in Industrial Technologies

Accurately predicting the Depth of Penetration (DOP) is essential for understanding the impact behavior of Functionally Graded Composite Materials (FGCMs) under high-velocity conditions. However, the nonlinear dynamic behavior governing FGCMs, coupled with the high cost and time-consuming nature of experimental tests, make precise DOP estimation a challenging task. This study proposes an efficient predictive framework that combines finite element (FE) simulations with artificial neural networks (ANNs) to overcome the constraints of conventional numerical and experimental approaches. This work focuses on applying the Whale Optimization Algorithm (WOA) and the Arithmetic Optimization Algorithm (AOA) as independent optimization strategies to improve the control parameters during the initiation phase, avoid convergence to local optima, and accelerate the learning process. Finite Element (FE) simulations were used to generate input-output datasets, which were used to train three ANN models: conventional back-propagation (BP), WOA-ANN, and AOA-ANN. The optimized ANNs accurately captured the nonlinear mapping between impact velocity ( <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" overflow="scroll"> <mml:mrow> <mml:msub> <mml:mi>V</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mi>m</mml:mi> <mml:mi>p</mml:mi> </mml:mrow> </mml:msub> </mml:mrow> </mml:math> ) and DOP. Results confirm that WOA-ANN model outperformed both AOA-ANN and BP models, achieving lower mean square error (MSE) and higher correlation coefficients. Moreover, the optimized ANNs provide reliable predictions with a limited number of FE simulations, considerably reducing computational effort while maintaining accuracy.

Abstract Cold metal transfer (CMT) is an advanced welding process developed by Fronius International, primarily designed for wire arc additive manufacturing (WAAM), a type of 3D metal printing. Combining the advantages of gas metal arc welding (GMAW) with precise droplet control, CMT is particularly well suited for additive manufacturing applications. Therefore, the use of stainless steel specifically for these applications requires knowledge of its microstructural and mechanical properties. In our recent study, a vertical wall was 3D-printed using 316L stainless steel via the CMT process. Three samples of bottom, middle, and top were taken of this as-deposited wall and were characterized by microstructural analyses by XRD, SEM, and OM, and extensive hardness testing via Vickers microhardness to evaluate the variation of their properties at the microscopic and macroscopic levels, respectively. The microstructure results indicate highly sensitive to the varying thermal gradients and cooling rates across the build direction. The bottom layers exhibit a fine, skeletal ferrite morphology due to efficient substrate heat dissipation, while heat accumulation in the middle layers results in a coarser, more equiaxed structure. Conversely, the top layers undergo a secondary refinement and exhibit increased residual ferrite content, as rapid cooling from ambient air limits the completion of the peritectic reaction. The hardness ranged from 181 to 200 Vickers hardness (HV), with a considerable increase at the top layers due to concentration of δ ferrite. These results provide a valuable contribution to the design of steel structures using 3D-printed stainless steel.

The present research introduces an electrochemical detection system constructed from chitosan-functionalized Ag–Cu-ferrite nanoparticles (Ag0.02Cu0.98Fe2O4@chitosan). This nanocomposite-based sensor provides an efficient approach for the targeted measurement of ascorbic acid (AA). Characterization using XRD and FTIR techniques confirmed the effective incorporation of chitosan into the Ag0.02Cu0.98Fe2O4 spinel framework. Using CV (cyclic voltammetry), the Ag0.02Cu0.98Fe2O4@chitosan-supported CPE electrode exhibited an oxidation peak at +0.28 V compared to +0.55 V for bare CPE in phosphate buffer (pH = 7), highlighting its superior electrocatalytic achievement. The current response exhibited a linear dependence with the square root of the scan rate (R2 = 0.9895), suggesting that the process is governed by diffusion. Sensitivities of 411.59 µA mM−1 cm2 and 327.31 µA mM−1 cm2 were achieved by the sensor over two distinct linear ranges, 100–300 µM and 300 µM–13 mM. With a high degree of certainty, the limit of detection (LOD) was recorded at 89 µM. The fabricated electrode displayed high selectivity within the presence of commonly used interferents (glucose, fructose, sucrose, NaCl, and KCl), as well as impressive reproducibility (RSD = 1.96%) and repeatability (RSD = 1.15%).It also achieved a superior AA recovery rate of up to 98% in a real sample, demonstrating its consistency and reliability. DFT computations confirmed AA's chemical reactivity through a moderate energy gap (ΔE = 5.42 eV) and identified reactive sites on oxygen atoms using molecular electrostatic potential (MESP) and Fukui function analysis (FIs). Additionally, Monte Carlo simulations (MCS) revealed a highly favorable adsorption energy of −26.65 kcal mol−1 for AA on the Ag0.02Cu0.98Fe2O4@chitosan-CPE surface, validating the strong affinity of the target analyte toward the modified electrode. Overall, the integration of chitosan into the Ag0.02Cu0.98Fe2O4 nanostructure significantly improved the electrocatalytic efficiency, selectivity, and stability of the sensor, making it an excellent option for practical applications in vitamin C detection.

This study presents a comparative investigation of the structural, morphological, and magnetic properties of nanostructured FeSi and FeSn alloys synthesized via the ball milling process. The alloys were analyzed at various milling times from 0 to 20 hours using X-ray diffraction (XRD), scanning electron microscopy (SEM), and vibrating sample magnetometer (VSM). XRD analysis confirmed the formation of FeSi and FeSn phases after 20 hours of milling, with characteristic peaks indicating structural evolution. SEM micrographs revealed significant morphological changes and particle refinement with increased milling time. Magnetic measurements highlighted contrasting behaviors between the two alloys. For FeSn, increased milling time enhanced coercivity, remanence magnetization, and squareness, accompanied by a reduction in saturation magnetization. Conversely, FeSi exhibited superparamagnetic behavior with notable saturation magnetization and significant coercivity (Hc). These findings emphasize the distinct magnetic and structural properties of FeSi and FeSn nanocrystalline alloys and their potential for tailored applications in advanced magnetic systems.

This study aims to develop a smart system for detecting internal defects in dielectric materials using capacitive sensing and 2D simulation through COMSOL. The system consists of a pair of electrodes placed above and below a composite sample. Defects such as air pockets or voids alter the material's relative permittivity, which directly affects the measured capacitance. A virtual 2D model was created featuring a rectangular sample with electrodes and an internal defect. Changes in capacitance due to this defect were calculated and visualized using Python as a heat map to locate defect areas. Several case studies were conducted with variations in defect size and position, and their effects on the electric field distribution and total capacitance were analyzed. The results demonstrate the effectiveness of the proposed approach in transforming electrical signals into visual data, paving the way for low-cost, intelligent non-destructive testing tools based on signal processing and smart imaging.

The detection of elliptical shapes in images plays a vital role in numerous computer vision applications, including medical imaging, industrial inspection and remote sensing. While traditional methods primarily focus on thin, well-defined elliptical contours using edge detection or curve fitting, real-world images often present thick, blurred, or partially occluded elliptical structures that challenge these techniques. In this work, we tackle the problem of detecting multiple thick elliptical regions by introducing a novel probabilistic approach based on Finite Mixture Models, where a new distribution, called Ellipse-Anchored Gaussian Distribution (EllAGD) is defined and which models the intensity distribution around elliptical shapes and incorporates a spatial thickness parameter that effectively captures thick contours. An Expectation-Maximization algorithm is used to estimate all the multiple ellipse parameters including spatial occupancy rate, center, axes, orientation angle and thickness. The proposed method is applied on synthetic and real images containing single or multiple thick elliptical structures where, it shows its ability to recover accurately all the structures parameters even with challenging image conditions.