NİKEL ESASLI INCONEL 718 SÜPER ALAŞIMININ İŞLENEBİLİRLİĞİNİN DEĞERLENDİRİLMESİ VE MAKİNE ÖĞRENİMİ KULLANILARAK ÇIKTI PARAMETRELERİNİN TAHMİN EDİLMESİ
Yükleniyor...
Dosyalar
Tarih
2024-06
Yazarlar
Dergi Başlığı
Dergi ISSN
Cilt Başlığı
Yayıncı
Erişim Hakkı
info:eu-repo/semantics/openAccess
Özet
Bu çalışmada, nikel esaslı Inconel 718 süper alaşımın frezelenmesinde soğutma/yağlama şartlarının ve kesme parametrelerinin güç tüketimi, yüzey pürüzlülüğü, kesme sıcaklığı, takım aşınması ve titreşim üzerindeki etkileri araştırılmıştır. Ayrıca işleme parametreleri ve deney sonuçları kullanılarak üç farklı regresyon tabanlı makine öğrenme modeli (Lineer Regresyon (LR), Karar Ağacı Sınıflandırıcısı (KAS) ve Destek Vektör Makinesi (DVM)) kullanılarak tahmin edilmiştir. Deney tasarımı Taguchi L27 dizinine göre hazırlanmıştır. İşleme parametresi olarak, üç farklı soğutma/yağlama şartı (kuru, hava ve MQL), üç farklı kesme hızı (40 m/dak, 60 m/dak ve 80 m/dak) ve üç farklı ilerleme miktarı (0,06 mm/dev, 0,09 mm/dev ve 0,12 mm/dev) seçilmiştir. Sonuç olarak en düşük güç tüketimi MQL yağlama şartında, 40 m/dak kesme hızında ve 0,06 mm/dev ilerleme miktarında 0,76 kW olmuştur. En düşük yüzey pürüzlülüğü ve titreşim MQL yağlama şartında, 80 m/dak kesme hızında ve 0,06 mm/dev ilerleme miktarında sırası ile 0,234 µm ve 1,91 m/s2 ölçülmüştür. Ayrıca en düşük kesme sıcaklığı hava soğutma şartında, 40 m/dak kesme hızında ve 0,06 mm/dev ilerleme miktarında 31 °C ölçülmüştür. Yapılan çalışmada kesme hızının, kesme sıcaklığının, ve titreşimin kesme hızının artması ile arttığı görülmüştür. Ancak artan kesme hızı yüzey pürüzlülüğünün azalmasına neden olmuştur. Artan ilerlme miktarı güç tüketiminin, yüzey pürüzlülüğünün, kesme sıcaklığının ve titreşimin artmasına sebep olmuştur. Kesme sıcaklığının azalması takım aşınması üzerinde olumlu bir etki göstermiştir. Ayrıca MQL yağlama şartı güç tüketimi, yüzey pürüzlülüğü ve titreşim üzerinde olumlu etki göstererek değerlerin azalmasına neden olmuştur. Lineer Regresyon, Karar Ağacı Sınıflandırıcısı ve Destek Vektör Makinesi kullanılarak güç tüketimi, yüzey pürüzlülüğü, kesme sıcaklığı ve titreşim için hata performans ölçütlerine ve elde edilen tahmini değerlere göre en iyi tahmini Destek Vektör Makinesi ile elde edilmiştir. Bunun için R-kare güç tüketimi, yüzey pürüzlülüğü, kesme sıcaklığı ve titreşim için sırası ile %94, %98, %93 ve %90 olmuştur.
In this study, the effects of cooling/lubrication conditions and cutting parameters on power consumption, surface roughness, cutting temperature, tool wear and vibration in milling nickel-based Inconel 718 super alloy were investigated. Additionally, processing parameters and experimental results were estimated using three different regression-based machine learning models (Linear Regression (LR), Decision Tree Classifier (DTC) and Support Vector Machine (SVM)). The experimental design was prepared according to the Taguchi L27 index. As machining parameters, three different cooling/lubrication conditions (dry, air and MQL), three different cutting speeds (40 m/min, 60 m/min and 80 m/min) and three different feed rates (0.06 mm/rev). , 0.09 mm/rev and 0.12 mm/rev) were selected. As a result, the lowest power consumption was measured at 0.76 kW at MQL lubrication, 40 m/min cutting speed and 0.06 mm/rev feed rate. As a result, the lowest power consumption was measured as 0.76 kW with MQL lubrication, at a cutting speed of 40 m/min and a feed rate of 0.06 mm/rev. The lowest surface roughness and vibration were measured as 0.234 µm and 1.91 m/s2, respectively, under MQL lubrication condition, 80 m/min cutting speed and 0.06 mm/rev feed rate. In addition, the lowest cutting temperature was measured as 31 °C under air cooling condition, at a cutting speed of 40 m/min and a feed rate of 0.06 mm/rev. In the study, it was observed that cutting speed, cutting temperature, and vibration increased with increasing cutting speed. However, increasing cutting speed caused the surface roughness to decrease. The increasing amount of feed caused an increase in power consumption, surface roughness, cutting temperature and vibration. The decrease in cutting temperature had a positive effect on tool wear. In addition, MQL lubrication condition had a positive effect on power consumption, surface roughness and vibration, causing the values to decrease. Using Linear Regression, Decision Tree Classifier and Support Vector Machine, the best prediction was obtained with the Support Vector Machine according to the error performance criteria and the estimated values obtained for power consumption, surface roughness, cutting temperature and vibration. For this, R-squared was 94%, 98%, 93% and 90% for power consumption, surface roughness, cutting temperature and vibration, respectively."
In this study, the effects of cooling/lubrication conditions and cutting parameters on power consumption, surface roughness, cutting temperature, tool wear and vibration in milling nickel-based Inconel 718 super alloy were investigated. Additionally, processing parameters and experimental results were estimated using three different regression-based machine learning models (Linear Regression (LR), Decision Tree Classifier (DTC) and Support Vector Machine (SVM)). The experimental design was prepared according to the Taguchi L27 index. As machining parameters, three different cooling/lubrication conditions (dry, air and MQL), three different cutting speeds (40 m/min, 60 m/min and 80 m/min) and three different feed rates (0.06 mm/rev). , 0.09 mm/rev and 0.12 mm/rev) were selected. As a result, the lowest power consumption was measured at 0.76 kW at MQL lubrication, 40 m/min cutting speed and 0.06 mm/rev feed rate. As a result, the lowest power consumption was measured as 0.76 kW with MQL lubrication, at a cutting speed of 40 m/min and a feed rate of 0.06 mm/rev. The lowest surface roughness and vibration were measured as 0.234 µm and 1.91 m/s2, respectively, under MQL lubrication condition, 80 m/min cutting speed and 0.06 mm/rev feed rate. In addition, the lowest cutting temperature was measured as 31 °C under air cooling condition, at a cutting speed of 40 m/min and a feed rate of 0.06 mm/rev. In the study, it was observed that cutting speed, cutting temperature, and vibration increased with increasing cutting speed. However, increasing cutting speed caused the surface roughness to decrease. The increasing amount of feed caused an increase in power consumption, surface roughness, cutting temperature and vibration. The decrease in cutting temperature had a positive effect on tool wear. In addition, MQL lubrication condition had a positive effect on power consumption, surface roughness and vibration, causing the values to decrease. Using Linear Regression, Decision Tree Classifier and Support Vector Machine, the best prediction was obtained with the Support Vector Machine according to the error performance criteria and the estimated values obtained for power consumption, surface roughness, cutting temperature and vibration. For this, R-squared was 94%, 98%, 93% and 90% for power consumption, surface roughness, cutting temperature and vibration, respectively."
Açıklama
Anahtar Kelimeler
Inconel 718 Süper Alaşım, Frezeleme, Soğutma/Yağlama Şartları, Güç Tüketimi, Yüzey Pürüzlülüğü, Kesme Sıcaklığı, Takım Aşınması, Titreşim, Makine Öğrenimi., Inconel 718 Super Alloy, Milling, Cooling/Lubrication Conditions, Power Consumption, Surface Roughness, Cutting Temperature, Tool Wear, Vibration, Machine Learning.
