Composite sandwich panels are widely used in aerospace, tra  ortation, and lightweight structural applications due to their high strength-to-weight and stiffness-to-weight ratios. One of the main challenges in designing these structures is improving their impact resistance and energy absorption capacity. This study investigates the low-velocity impact response of a composite sandwich panel and evaluates the effects of structural modifications on its energy absorption and specific energy absorption (SEA). To this end, a numerical model of the panel was developed using Abaqus and validated against reference results for hollow and polymethacrylimide (PMI) foam-filled panels at impact energies of 25 and 100 J. The prediction errors ranged from 4.02% to 6.91% for the hollow panel and from 7.19% to 9.12% for the PMI foam-filled panel, indicating acceptable agreement with the reference results. Following model validation, the effects of increasing the number of face-sheet plies from six to eight, replacing the outermost face-sheet plies with Kevlar fibers, and substituting the PMI foam core with polyurethane (PUR) foam were investigated. The results revealed that the effect of increasing the number of face-sheet plies was dependent on the impact energy. At lower impact energies, adding plies did not necessarily improve the impact performance, whereas at higher energies, the additional plies played a more significant role in energy absorption. Incorporating Kevlar fibers into the outer face-sheet plies improved energy absorption at all investigated impact energy levels and resulted in the highest SEA among the configurations considered. Furthermore, the PMI foam core exhibited higher total energy absorption, while the lower mass of the PUR foam core resulted in higher SEA under certain conditions. Overall, the impact response of composite sandwich panels was found to depend on the number and material composition of the face-sheet plies, the core material, and the total structural mass. Therefore, evaluating both total absorbed energy and SEA provides a more comprehensive basis for designing and optimizing lightweight sandwich panels with enhanced impact resistance. Keywords: Low-velocity impact, composite sandwich panel, energy absorption, honeycomb core, numerical model.

      Abstract: With the increasing adoption of photovoltaic systems and the growing importance of improving reliability in renewable energy generation systems, rapid and accurate fault diagnosis has become one of the major challenges in this field. The present study was conducted with the aim of fault detection and >  

   Abstract: Among the various types of faults in electrical machines, air-gap eccentricity is considered one of the most important and complex mechanical faults. This fault occurs when the air gap between the rotor and stator is not uniform around the circumference of the machine. Under ideal conditions, the geometric axes of the rotor and stator are perfectly aligned, resulting in a constant air-gap length. However, manufacturing errors, improper installation or assembly, bearing wear, shaft bending, rotor unbalance, and deformation of mechanical components can disrupt this symmetry. Consequently, the magnetic field becomes asymmetrically distributed around the machine, generating unbalanced magnetic forces that directly affect both the electromagnetic and mechanical performance of the machine. Since the finite element analysis (FEA) method enables accurate modeling of the actual machine geometry, nonlinear magnetic material properties, core saturation, slot effects, and local variations in the air gap, this thesis employs this analytical method to investigate and analyze the behavior of a switched reluctance machine under eccentricity faults. The results obtained from finite element simulations demonstrate that rotor eccentricity causes an asymmetric distribution of magnetic flux density around the air gap, which consequently affects the air-gap flux, phase currents, phase inductances, and ultimately the developed torque. Therefore, analysis of these parameters can be used to detect and estimate the degree of rotor eccentricity. Subsequently, based on the results obtained from finite element analysis, a multilayer perceptron (MLP) neural network was developed to estimate the degree of rotor eccentricity in switched reluctance motors (SRMs) using the electromagnetic and performance parameters of the motor. The dataset consists of 16 samples obtained from two three-phase and four-phase motors with different eccentricity levels, ranging from zero to approximately 2 mm. The network inputs include the average phase flux linkage, average output torque, torque ripple percentage, and average phase current, while the network output is the rotor eccentricity expressed in millimeters. Following normalization, the data were used to train the neural network. Upon completion of the training process, the performance of the network was evaluated using metrics such as mean squared error (MSE), mean absolute error (MAE), and coefficient of determination (R²) to assess the model's ability to estimate the rotor eccentricity level and reproduce the underlying patterns in the data. Keywords: Switched Reluctance Motor, Fault, Deep Learning, Neural Network