Papers by Keyword: Model

Abstract: The use of out-of-furnace desulphurization of cast iron and various dispersed desulfurizing reagents is due to the desire to ensure the most complete removal of cast iron sulfur in the shortest period of time. The actual results of the industrial application of out-of-furnace desulfurization indicate that the practical results and application rates in a number of cases are not stable enough and are far from possible and expected. The studies were carried out on calculated and "cold" transparent physical models. Magnesium, lime, and calcium carbide were evaluated as desulfurizing reagents. Based on the actual results of physical modeling and subsequent calculations, an improved expression was formulated for determining the length of a gas jet in a liquid (L_str ) - the depth of the jet immersion, depending on the parameters of injection through a submerged lance. The processes of interaction between gas and solid phases in the near-lance zone during ladle desulphurization have been studied. It is shown that during the injection desulfurization of cast iron, the gas component of the flow stops its directional movement in the melt for up to 80 mm (practically 50–60 mm), solid particles continue to move in the bubble and hit the surface of this cavity. To assess the further movement of the particle through the "gas cavity-melt" boundary, the depth of penetration of particles into liquid iron was calculated. The motion of a particle in a melt can be described by an equation that is arranged for the conditions of vertical motion of a particle from top to bottom with a given initial velocity up to the complete stop of the particle. Nomograms are given to determine the specified parameters. Recommendations are given on the parameters of injection of magnesium and ground lime.

Abstract: The article summarizes the results of comprehensive theoretical and experimental studies of ice fracture under shock and explosive loads. Artificial ice and freshwater river ice were considered as objects of research. The results of full-scale underwater explosive tests are presented. Post-explosion analysis of the crushing of a 130-day ice sheet, including the morphology of destruction and diameter and state of the ice edge were obtained. The results of a five-layer ice target impacted by a low-velocity striker showed that a brittle fracture mechanism was dominant. A phenomenological model of ice destruction is briefly described. The model was a complex one of continuum mechanics and was based on fundamental conservation laws. The ice failure concept was based on a deterministic approach and the combined use of several failure criteria. The finite-element Lagrangian method contained a new method for isolating the discontinuity surfaces of materials. The calculations were carried out using the noncommercial software package Udar.Os.1. The impact of an ice cylinder on a rigid wall (aluminum plate) was simulated. Good agreement was obtained in terms of the morphology of the fracture and the velocity of the fracture wave. The contact surface algorithm was illustrated, which helped save computational time when modeling some problems of perforation and penetration, including the detonation process. In the numerical experiment, ice without phase transitions with averaged mechanical properties was considered. The impact response of the ice blocks to the shock and explosive load was simulated. The perforation of structures consisting of ice cubes and thin steel plates above and on them is simulated. Deep penetration of the steel sphere into an ice block and an ice block protected by a metal plate was simulated. Using numerical modeling, the location of explosive substances for the most effective fracture of thick ice was determined.

Abstract: This scientific work presents the development of a computer-simulation model for particle filling in three-dimensional space based on molecular dynamics methods. The Lennard-Jones potential was used to simulate interactions between particles, and the equations of motion were integrated using the Velocity Verlet algorithm. The model incorporates periodic boundary conditions (PBC), ensuring accurate representation of an infinite system without boundary effects. The simulation results confirm the system's energy stability: the total energy remains virtually unchanged throughout the simulation, indicating the correctness of numerical integration. Fluctuations in kinetic and potential energies demonstrate normal system dynamics, where energy is redistributed among particles through interactions. An analysis of the spatial distribution of particles revealed that the system remains in a liquid state, with no signs of solid structure formation or particle aggregation. Notably, the developed model enables the simulation of complex physical processes such as dense structure formation, particle transport, and self-packing. The obtained results highlight the efficiency of the molecular dynamics method for analyzing granular and particulate media, as well as for studying the physical properties of multi-particle systems. The model can be utilized to optimize technological processes related to material transportation, packaging, and storage, as well as for research into nanomaterials and composites.

Abstract: Financial fraud remains a persistent challenge across various domains, particularly in public-sector financial operations, threatening the integrity and transparency of financial statements while eroding public trust. This highlights the need for continued advancement of fraud detection mechanisms to keep up with the ever-evolving fraud tactics. ML algorithms have proven to be one of the most successful methods for analysing large financial datasets to detect fraudulent patterns. This paper reviews the application of ML to detect fraud in financial transactions using ML-based algorithms, namely K-means, Support Vector Machine, Decision Trees, Naive Bayes, and Deep Learning, in fraud detection, analysing their use cases and effectiveness as reported in the literature. Additionally, the study experimentally compares the performance of a Convolutional Neural Network (CNN) model against a Logistic Regression model, with the CNN achieving an impressive 90% accuracy, outperforming Logistic Regression in fraud detection. The paper further investigates the financial features and indicators most relevant to fraud detection and explores the challenges and opportunities posed by large volumes of financial transactions. By addressing these areas, the study aims to provide insights into enhancing fraud detection mechanisms and strengthening the security and integrity of financial transactions in today's digital ecosystem, including government institutions.

Abstract: The development of remote sensing technology based on cloud computing is an important concern, especially in providing convenience in processing and storing images for various importance. This can be utilized in hydrographic activities in terms of determining depth values using the Satellite Derived Bathymetry (SDB) method. This research focuses on the utilization of Google Earth Engine (GEE) technology in obtaining depth values using the SDB method with Stumpf and Lyzenga algorithms. The area used in this research is on the island of Bangka Belitung which has different characteristics of waters such water depth, water clarity, and water bottom cover. The results of the research conducted were evaluated with the Indonesian National Standard (SNI) 8202 in 2015 regarding the accuracy of the base map. Depth extraction using 2,662 depth sample data. The results of this study show that the Stumpf model provides a better accuracy value compared to the Lyzenga model at 0 - 10 meters. But at 10-25 meters the Lyzenga model produces a better accuracy value than the Stumpf model. The best correlation value was obtained with the Lyzenga model, which is equal to 0.841 while the correlation value of the Stumpf model is 0.753. The best RMSE was obtained at 1.7 m with a depth range of 0 - 5 m with the Stumpf method. The results of the accuracy evaluation of the two empirical bathymetry models pass the requirements of SNI 8202 of 2015 regarding the Accuracy of the Base Map for making Lingkungan Pantai Indonesia (LPI) and Lingkungan Laut Nasional (LLN) maps at a scale of 1: 50,000 with a 20 m contour interval.

Abstract: Riceberry rice has a dark purple color; and a high content of antioxidants, which could affect the digestion behaviors and its application. This study is aimed to analyze the starch digestion rate and predict the bio-accessibility of polyphenols in various modified Riceberry flours during the in vitro digestive process. It also discussed the relationship between the rate of digestion and polyphenol release, which provided basic information about the digestion behavior of Riceberry flour. Seven rice flour samples were used for this study, which included six physically treated flours: annealed flour (AF), heat moisture-treated flour (HMT), pregelatinized flour (Pregel), ultra-sonicated flour (US), wet microwave-treated flour (Wet), dry microwave treated flour (Dry), and untreated (control sample). The obtained results showed that, compared with the control sample, the digestion rate of the Pregel sample was higher, while the others had lower values. However, the Pregel sample showed the second highest rank of bio-accessible polyphenol during digestion after the US sample. While the HMT sample presented the lowest rate of starch digestion and release of bioactive compounds. This investigation also used an artificial neural network (ANN) to forecast the starch digestion and polyphenol bio-accessibility of rice flours. During digestion, the ANN model demonstrated a high capacity to predict the polyphenol bio-accessibility and starch hydrolysis percentage. There was a goodness of fit between the ANN-predicted and the actual values (R² >0.95). The importance of the bioavailability and bio-accessibility analysis indicates the functional potential that flour can have, which could be predicted effectively by applying modern techniques such as the ANN model. Moreover, it was also concluded that the digestive tract readily absorbs released polyphenol compounds in rice flour, which also influences the rate of starch hydrolysis. However, the impact could vary depending on the flour’s starch fraction content and the polyphenol activity, which is a topic for future investigation. The high antioxidant content and low digestion rate of flour could be highly promising functional materials for application in the food industry.

Abstract: Improved understanding of friction during cold rolling is crucial to further optimize the rolling process, to accurate analyse cold rolling defects and to increase model accuracy enabling an improved mill setup during industrial operation. Classical slab rolling models make use of the Coulomb friction law, assuming a constant coefficient of friction in the roll bite. In the last decades, mixed-lubrication models have been developed that explicitly take the lubricant action into account. These models have greatly increased the understanding of factors that influence friction during cold rolling, but quantitatively the model results should still be further improved before such models can be used as an online tool for setting up the cold rolling mill. This article describes a mixed-lubrication model to simulate cold rolling of low-carbon steel. Especially the tribological core of the model is extended and improved compared to state-of-the-art models. Friction mechanisms now also include a viscous shear stress and ploughing friction. The quantification of viscous shear stress was reported in a previous work [1], this work focuses on the quantification of ploughing friction. Material Point Method (MPM) simulations were carried out to determine the work piece strain-hardening and strainrate-hardening under a ploughing indenter. These simulations result in an ‘Surface Ploughing Resistance’ and finally in a quantification of the contribution of ploughing friction to the overall friction in the roll bite. The description of the various friction mechanisms (ploughing, adhesive and viscous shear) is implemented in the mixed-lubrication model. This article concludes by presenting typical results of the developed model. One of the main conclusions is that the contribution of ploughing friction in a cold rolling process cannot a priori be neglected.

Abstract: The article presents a simulation model for determining the required area of easily removable structures to protect against progressive collapse. The simulation model allows you to calculate the area of easily removable structures depending on the input parameters, to obtain the dynamics of the change in the parameters of the combustible medium depending on the change in the properties of the combustible gases that are in the room.

Abstract: This study investigates the corrosion inhibition performance of Anthocleista grandiflora leaf (AGL) extract on carbon steel in seawater, considering the effects of temperature, immersion time, and inhibitor concentration. Predictive modeling, adsorption behavior, and the kinetics and thermodynamics of the inhibition process were examined. The weight loss technique,characterization techniques combined with response surface methodology (RSM), revealed that the AGL extract follows the Langmuir adsorption model, exhibiting physical adsorption with ΔG values between −16.24 to −15.49kJ/mol, indicating spontaneous and endothermic inhibition. The thermodynamic parameters entropy (−198.87 to −52.58 J/mol), enthalpy (20.42 to 53.42 kJ/mol), and activation energy (13.68 to 56.32 kJ/mol further support this. The corrosion reaction follows first-order kinetics, with the half-life decreasing as the rate constant and extract concentration increase.The SEM images revealed that the AGL extract formed a protective surface layer on the mild steel, effectively preventing pitting. This protective effect became more pronounced as the concentration of the extract increased. RSM optimization identified optimal conditions for maximum inhibition efficiency (98.70%) and corrosion rate (0.058 mm/y) at 800 ppm, 303 K, and 45 days, with a prediction accuracy of 95%, making it suitable for application in the oil and gas industry.

Abstract: High-entropy alloys (HEAs) have excellent properties that are being explored for potential applications in many engineering fields. Their excellent properties strongly depend on their phases. The vastness of alloy compositions that can be synthesized makes it extremely challenging to experimentally investigate all the possible HEA types. To mitigate these challenges, more efficient and systematic computational techniques can be applied to the existing experimental data to accelerate HEA design and discovery. Therefore, this study developed three soft computing classification models based on artificial neural network, k-nearest neighbor (kNN), and support vector machine (SVM) to classify solid solution, amorphous and intermetallic phases in HEAs. Empirical studies showed that hyperparameter optimization improved classification accuracies of the classifiers with kNN (92%) outperforming ANN (86%) and SVM (90%) using all five predictive features. Feature selection did not improve the classification accuracy of any of the model. This studied demonstrated the importance of applying soft computing techniques and hyperparameter optimization for enhancing the classification accuracies of models to predict the phases in HEAs.