Artificial neural network approach was used to predict the thicknesses of total (FeB+Fe2B), FeB and Fe2B borides layers of AISI 1020, AISI 1060, and AISI 4140 steels. Boronizing heat treatment was conducted in a solid medium comprising of EKabor®2 powders at 840–960 ˚C at 40 ˚C intervals for 2, 4, 6, and 8 hours. Optical microscope analysis of the borided layer revealed the saw-tooth (columnar) and planar morphology. The depth of the total (FeB+Fe2B), FeB and Fe2B boride layers was accurately predicted. For total boride layers generated by the artificial neural network model, the average error varied between 0.04 and 7.64 µm. Micro hardness values increased by 423% in AISI 1020, 336% in AISI 1060, and 411% in AISI 41040 after the boronizing process.
In this study, the effects of atmospheric pressure cold and hot plasma treatment on the adhesive bond strength of 5049 aluminum alloy (5049) and carbon fiber reinforced polymer composite (CFRP) were investigated. For this purpose, plasma treatments were applied to sample surfaces prepared in accordance with the standard dimensions. The experimental design was created with a Central Composite Design using the Response Surface Method (RSM) with 3 different application distances and 3 different numbers of passes as plasma parameters. The contact angles of samples to which atmospheric pressure cold and hot plasma were applied and which were not plasma treated were measured. Depending on the experimental design, the adhesive strengths of the single-lap joint samples were determined by tensile test after the applied plasma treatments. As a result of atmospheric pressure cold plasma application, the average contact angle values decreased by 13.68% on the 5049 surface and 11.86% on the CFRP surface. As a result of hot plasma application, the average contact angle values decreased by 26.04% on the 5049 surface and 30.08% on the CFRP surface. Shear stress results of single-lap joints showed an increase of 5.83% after atmospheric pressure cold plasma and 7.97% after hot plasma.
In this study, the effects of atmospheric pressure cold plasma treatment on surface contact angles and adhesive joint performance of aluminum 7075-T6 alloy and carbon fiber reinforced polymer composites were investigated using experimental, numerical, and statistical methods. A full factorial experimental design was applied to evaluate three plasma parameters: application distance, number of passes, and application speed. Contact angle measurements were conducted to assess surface activation, and single lap adhesive joints were fabricated and tested to determine lap shear strength. The results were analyzed using analysis of variance to identify significant parameters and their contribution ratios. Process–output relationships were modeled using response surface methodology. Finite element analyses with LS-DYNA were performed to investigate stress distributions and mechanical behavior of the adhesive joints. Multi-objective optimization was carried out using the non-dominated sorting genetic algorithm II to minimize contact angles and maximize lap shear strength. Application distance was found to be the most influential parameter. The lowest contact angles and highest lap shear strength were achieved at high number of passes and low application distance and speed. Numerical results showed good agreement with experiments, and a Random Forest model was used to improve prediction accuracy within the optimization framework.
31 July 2026 in Balıkesir Üniversitesi Fen Bilimleri Enstitüsü Dergisi
In this study, the surface roughness values obtained as a result of sanding the surfaces of S235 and polypropylene (PP) materials with different sanding parameters were investigated. Sanding size (P60, P100, P180, and P320) and sanding time (15, 30, 45, and 60 seconds) were used as sanding parameters. The effect ratios on material roughness were determined by analysis of variance. It was concluded that sanding size was the most effective parameter on surface roughness. Mathematical models and artificial neural networks were created for prediction. Surface roughness decreased with increasing sanding size number and sanding time. In the applied parameters, it was measured that there was a decrease in surface roughness of over 80% for both materials. Optimum parameters were determined based on the criterion of minimizing the surface roughness (Ra) value. As a result of the optimisation with Response Surface Method (RSM), a 60-second sanding process with P320 sandpaper was determined as the optimum condition for achieving the minimum Ra value.