The present study aims to investigate the effects of cutting parameters (cutting speed, Vc: 60–90–120 m/min; feed rate, f: 0.055–0.085–0.115 mm/rev), cutting tool coatings (CVD: TiN/TiCN/Al₂O₃ and PVD: TiAlN), and machining conditions (dry, air, and MQL) on cutting force (Fc), specific energy consumption (SEC), surface roughness (Ra), cutting temperature (T), and tool wear (Vb) during the milling of Ti6Al4V alloy. As a result, it was observed that all machining tests conducted with the Al₂O₃-coated cutting tool showed improvements of 4.7%, 10.75%, 3.8%, and 6.3% in Fc, SEC, Ra, and T, respectively, compared to the tests performed with the TiAlN-coated cutting tool. Under dry machining conditions, the average Fc, SEC, Ra, and T values were 302.82 N, 4.88 j/mm³, 0.653 µm, and 241.06 °C, respectively. Compared to dry machining conditions, the air and MQL machining conditions demonstrated improvements in the average Fc by 5.15% and 6.3%, SEC by 10.27% and 17.79%, Ra by 6.23% and 11.17%, and T by 8.9% and 19.68%, respectively. The lowest Fc and Ra values for the Al₂O₃-coated cutting tool were measured at 228.33 N and 0.402 µm, respectively, under the MQL machining condition, at a cutting speed of 120 m/min and a feed rate of 0.055 mm/rev. The lowest SEC value (2.694 J/mm³) was also obtained using the Al₂O₃-coated tool under MQL conditions at a cutting speed of 120 m/min and a feed rate of 0.115 mm/rev. Similarly, the lowest cutting temperature (129 °C) was achieved with the Al₂O₃-coated tool under MQL conditions at a cutting speed of 60 m/min and a feed rate of 0.055 mm/rev. The wear performance of the Al₂O₃-coated cutting tool was observed to be superior to that of the TiAlN-coated tool.

In this study, the drilling of an Al 6082-T6 alloy and the effects of cutting tool coating and cutting parameters on surface roughness, cutting temperature, hole diameter, circularity, and cylindrical variations was investigated. In addition, the prediction accuracy of Taguchi, artificial neural networks (ANNs), and adaptive neuro-fuzzy inference system (ANFIS) methods was compared using both experimental results and Signal/Noise (S/N) ratios derived from the experimental results. The experimental design was prepared according to Taguchi L27 orthogonal indexing. As a result, it was observed that increasing the cutting speed and feed rate increases the cutting temperature hole error, circularity error and cylindricity error. Increasing the cutting speed positively affected the surface roughness, while increasing the feed rate led to an increase in the surface roughness. The lowest surface roughness, cutting temperature, hole diameter error and hole circularity error values were measured for the uncoated cutting tool. The minimum cylindricity variation was measured for drilling with TiAlN-coated cutting tools. The optimum cutting parameters were A1B1C3 (Uncoated, 0.11 mm/rev, 200 m/min) for surface roughness, A1B1C1 (Uncoated, 0.11 mm/rev, 120 m/min) for cutting temperature, hole error, circularity error and cylindricity error. In the estimation of the output parameters with Taguchi, ANNs and ANFIS, it was observed that the estimates made by converting the experimental values into S/N ratios were more accurate than the estimates made with the experimental results. The reliability coefficient and prediction ability of the ANN model were found to be higher than Taguchi and ANFIS models in estimating the output parameters.

This paper focuses on the implementation of the Whisper architecture to create an automatic speech recognition (ASR) system optimized for the Turkish language, which is considered a low-resource language in terms of speech recognition technologies. Whisper is a transformer-based model known for its high performance across numerous languages. However, its performance in Turkish, a language with unique linguistic features and limited labeled data, has yet to be fully explored. To address this, we conducted a series of experiments using five different Turkish speech datasets to assess the model’s baseline performance. Initial evaluations revealed a range of word error rates (WERs) between 4.3% and 14.2%, reflecting the challenges posed by Turkish. To improve these results, we applied the low-rank adaptation (LoRA) technique, which is designed to fine-tune large-scale models efficiently by introducing a reduced set of trainable parameters. After fine-tuning, significant performance improvements were observed, with WER reductions of up to 52.38%. This study demonstrates that fine-tuned Whisper models can be successfully adapted for Turkish, resulting in a robust and accurate end-to-end ASR system. This research highlights the applicability of Whisper in low-resource languages and provides insights into the challenges of and strategies for improving speech recognition performance in Turkish.