DSpace Ain Temouchent Belhadj Bouchaib

Item

Simulation de la propagation de fissures Dans les structures sous chargement cyclique

(2026) MECELLEM MOHAMED ABDELLAH; ELRARBI SARA; CHAMA Mourad

Les bimatériaux sont largement utilisés dans différences branches de la technologie comme la mécanique, l’électronique et l’électrotechnique parce qu’ils sont constitués de deux matériaux dont les propriétés physiques et mécaniques sont totalement différentes. Le comportement en ruptures des matériaux hétérogènes nécessite des connaissances approfondies et le domaine de recherche reste complexe et ouvert. Le travail ainsi présenté a pour objectif d’analyser numériquement par la méthode éléments finis la propagation d’une fissure initiée dans la céramique et au voisinage proche de l'interface sous l’effet d’un chargement cyclique. L'étude porte sur l'analyse d’un modèle géométrique bidimensionnel, est une poutre en flexion, elle est composée de deux matériaux, alumine contient de teneur d’époxy avec et sans gradient au voisinage proche de l’interface. Une étude comparative a été faite en comportements élastiques. Plusieurs facteurs ont été mis en évidence tels que la résistance de la fatigue de la fissure et sa déviation dans la céramique. Abstract Bimaterials are widely used in various branches of technology, such as mechanics, electronics, and electrical engineering, because they are composed of two materials with completely different physical and mechanical properties. The fracture behavior of heterogeneous materials requires in-depth knowledge, and the research field remains complex and open. The work presented here aims to numerically analyze, using the finite element method, the propagation of a crack initiated in the ceramic and in the immediate vicinity of the interface under the effect of cyclic loading. The study focuses on the analysis of a two-dimensional geometric model, a beam in bending, composed of two materials: alumina containing epoxy with and without a gradient in the immediate vicinity of the interface. A comparative study was conducted on elastic behavior. Several factors were highlighted, such as the fatigue resistance of the crack and its deflection in the ceramic..

Item

PHOTOGRAMMETRY: EVOLUTION, CURRENT STATE, AND FUTURE TRAJECTORIES WITH A FOCUS ON BIM INTEGRATION

(2026) BEN AISSA RAHAL NOUR EL HOUDA MOKHTARIA; ZENASNI FATIMA ZOHRA; KADDOUR Hakim

BIM (Building Information Modeling) is considered one of the most advanced systems currently available for managing and monitoring engineering projects across all their phases and typologies. The system derives its strength from cutting-edge field data acquisition and processing technologies, such as LiDAR and XR. However, most of these technologies remain prohibitively expensive, particularly for small-scale projects. This is where photogrammetry emerges as a relatively cost-effective alternative. This thesis aims to clarify this technology and assess its potential to support BIM workflows, by examining its various types, stages of development, and underlying mathematical models, while identifying the advantages of each approach. To this end, field imagery was processed using three different photogrammetric methods SfM, NeRF, and Gaussian Splatting employing various software packages and computer systems with differing specifications. The results revealed no significant difference in the field image acquisition process; however, the data processing pipelines and resulting outputs varied considerably. While SfM leads as the foundational method for all photogrammetric approaches, offering the highest geometric accuracy in 3D models, Gaussian Splatting excels in the visual realism of its outputs, albeit at the cost of larger file sizes. NeRF ،on the other hand, produces AI-generated models with impressive visual appearance, but lacks true geometric structure, rendering it unsuitable for precise engineering measurements.

Item

Prédiction des propriétés à l’état durci du béton léger à l’aide de l’intelligence artificielle

(2026) BACHIR RAYANE ABDELAZIZ; HADJOUTI SAID; DOUNANE Nawal

Lightweight concrete has increasingly gained importance in civil engineering due to its numerous advantages, particularly the reduction of structural self-weight and the improvement of thermal insulation. However, its complex composition limits the effectiveness of conventional experimental methods. These methods face several constraints, including high costs, long testing durations, and the inability to explore all possible concrete formulations. In this context, this study develops and compares four machine learning models : Artificial Neural Networks (ANN), Random Forest (RF), Support Vector Machines (SVM), and XGBoost. The main objective is to simultaneously predict four key properties of lightweight concrete : compressive strength (CS), tensile strength (TS), dry density (DD), and modulus of elasticity (MOE). The dataset was compiled from international scientific literature and includes twelve input variables. The results highlight the superior performance of the XGBoost model, which achieved a coefficient of determination of R2 = 0.97 for predicting both compressive strength and dry density. The Random Forest model demonstrated the best performance for tensile strength prediction, with an R2 = 0.97. In contrast, the SVM model exhibited the lowest performance among the studied models, with an average R2 value of 0.84. A comparison with findings reported in the literature confirms that the machine learning models investigated outperform traditional prediction approaches. Therefore, these models offer promising opportunities for the optimization and design of new lightweight concrete mixtures with greater efficiency and accuracy.

Item

Évaluation de l’efficacité des biofertilisants naturels sur sols pauvres en matière organique

(2026) Belmokhtar Hena Soheir; Benadjila Rania; YAZIT SIDI-MOHAMMED

Soil fertility is a key factor in ensuring sustainable agricultural production. In this context, the valorization of organic waste represents an environmentally friendly alternative for improving soil properties while reducing the environmental impact of waste. The objective of this study was to evaluate the effect of coffee grounds and potato peels, used as organic amendments, on the physicochemical properties of soil and on the growth of lettuce (Lactuca sativa L.). To achieve this objective, four treatments were established: a control soil (T0), soil amended with coffee grounds (T1), soil amended with potato peels (T2), and soil amended with a mixture of coffee grounds and potato peels (T3). Plant growth parameters (plant height, number of leaves, leaf length, fresh biomass, and dry biomass) as well as selected soil chemical properties (pH, organic matter, and electrical conductivity) were evaluated after composting and after cultivation. The results showed that the organic amendments improved soil properties and promoted lettuce growth. Treatment T2 recorded the best performance, with a final plant height of 16 cm, a leaf length of 15 cm, and the highest fresh (9.53 g) and dry (0.55 g) biomass values. Soil analyses also revealed an increase in organic matter content in the amended treatments, while soil pH remained close to neutrality. An increase in electrical conductivity was observed after cultivation, reflecting the gradual release of mineral nutrients during the decomposition of the organic amendments. These findings highlight the agronomic potential of coffee grounds and potato peels as organic amendments capable of improving soil fertility and enhancing lettuce growth, thereby contributing to more sustainable agricultural practices.

Item

Prédiction des propriétés à l’état durci du béton léger à l’aide de l’intelligence artificielle

(2026) BACHIR RAYANE ABDELAZIZ; HADJOUTI SAID; DOUNANE Nawal

Lightweight concrete has increasingly gained importance in civil engineering due to its numerous advantages, particularly the reduction of structural self-weight and the improvement of thermal insulation. However, its complex composition limits the effectiveness of conventional experimental methods. These methods face several constraints, including high costs, long testing durations, and the inability to explore all possible concrete formulations. In this context, this study develops and compares four machine learning models : Artificial Neural Networks (ANN), Random Forest (RF), Support Vector Machines (SVM), and XGBoost. The main objective is to simultaneously predict four key properties of lightweight concrete : compressive strength (CS), tensile strength (TS), dry density (DD), and modulus of elasticity (MOE). The dataset was compiled from international scientific literature and includes twelve input variables. The results highlight the superior performance of the XGBoost model, which achieved a coefficient of determination of R2 = 0.97 for predicting both compressive strength and dry density. The Random Forest model demonstrated the best performance for tensile strength prediction, with an R2 = 0.97. In contrast, the SVM model exhibited the lowest performance among the studied models, with an average R2 value of 0.84. A comparison with findings reported in the literature confirms that the machine learning models investigated outperform traditional prediction approaches. Therefore, these models offer promising opportunities for the optimization and design of new lightweight concrete mixtures with greater efficiency and accuracy.

Communautés dans DSpace

Dépôts récents