Experimental investigation of wear behavior in Al2O3-reinforced glass fiber composites and comparative analysis of artificial neural network and machine learning models

Raşit Koray Ergün; İsmail Bayar; Hüseyin Köse

Article Details

Experimental investigation of wear behavior in Al2O3-reinforced glass fiber composites and comparative analysis of artificial neural network and machine learning models

Authors:

Raşit Koray Ergün | İsmail Bayar | Hüseyin Köse

Publisher:

Nigde Omer Halisdemir University

Main Topics:

Makine Öğrenme (Diğer) | Triboloji | Kompozit ve Hibrit Malzemeler

Journal:

Niğde Ömer Halisdemir Üniversitesi Mühendislik Bilimleri Dergisi Vol.14 Issue 4, pp.1571-1581

Document Type:

Article

Publication Date:

10/2025

Document Language:

English

Document ID (Org No.):

2001360305

pISSN:

2564-6605

Keywords:

Abstract:

 This study experimentally investigates the effects of adding different amounts 1-5 wt. of Al2O3 particles on the wear behavior of glass fiber-reinforced epoxy composites to improve their tribological performance. Composite laminates produced using the hand-lay up method were subjected to wear tests using a ball-on-disc test setup under dry sliding conditions. Among all tested compositions, the composite containing 3 wt. Al2O3 exhibited the highest wear resistance. Compared to the neat composite, the specific wear rate was reduced by up to 70. In contrast, 4 and 5 Al2O3 additions resulted in a decrease in wear resistance due to particle agglomeration. While the highest specific wear rate was 26010⁶ mm³Nm, this value decreased to 8010⁶ mm³Nm in the 3 added sample. Furthermore, wear rate predictions were performed using models such as artificial neural network and different machine learning regressors. Random Forest 17.62, Ridge regressor 18.46 and artificial neural network 19.92 achieved the lowest MAPE values, indicating strong predictive performance for Al2O3-reinforced glass fiber composites. The artificial neural network model optimized with grid search achieved a mean squared error of 0.90 and a coefficient of determination of 0.92, while the random forest regressor demonstrated strong generalization with a coefficient of determination of 0.91. The results demonstrated the critical roles of both particle ratio and data-driven models in wear performance analysis.

 This study experimentally investigates the effects of adding different amounts 1-5 wt. of Al2O3 particles on the wear behavior of glass fiber-reinforced epoxy composites to improve their tribological performance. Composite laminates produced using the hand-lay up method were subjected to wear tests using a ball-on-disc test setup under dry sliding conditions. Among all tested compositions, the composite containing 3 wt. Al2O3 exhibited the highest wear resistance. Compared to the neat composite, the specific wear rate was reduced by up to 70. In contrast, 4 and 5 Al2O3 additions resulted in a decrease in wear resistance due to particle agglomeration. While the highest specific wear rate was 26010⁶ mm³Nm, this value decreased to 8010⁶ mm³Nm in the 3 added sample. Furthermore, wear rate predictions were performed using models such as artificial neural network and different machine learning regressors. Random Forest 17.62, Ridge regressor 18.46 and artificial neural network 19.92 achieved the lowest MAPE values, indicating strong predictive performance for Al2O3-reinforced glass fiber composites. The artificial neural network model optimized with grid search achieved a mean squared error of 0.90 and a coefficient of determination of 0.92, while the random forest regressor demonstrated strong generalization with a coefficient of determination of 0.91. The results demonstrated the critical roles of both particle ratio and data-driven models in wear performance analysis.

 This study experimentally investigates the effects of adding different amounts 1-5 wt. of Al2O3 particles on the wear behavior of glass fiber-reinforced epoxy composites to improve their tribological performance. Composite laminates produced using the hand-lay up method were subjected to wear tests using a ball-on-disc test setup under dry sliding conditions. Among all tested compositions, the composite containing 3 wt. Al2O3 exhibited the highest wear resistance. Compared to the neat composite, the specific wear rate was reduced by up to 70. In contrast, 4 and 5 Al2O3 additions resulted in a decrease in wear resistance due to particle agglomeration. While the highest specific wear rate was 26010⁶ mm³Nm, this value decreased to 8010⁶ mm³Nm in the 3 added sample. Furthermore, wear rate predictions were performed using models such as artificial neural network and different machine learning regressors. Random Forest 17.62, Ridge regressor 18.46 and artificial neural network 19.92 achieved the lowest MAPE values, indicating strong predictive performance for Al2O3-reinforced glass fiber composites. The artificial neural network model optimized with grid search achieved a mean squared error of 0.90 and a coefficient of determination of 0.92, while the random forest regressor demonstrated strong generalization with a coefficient of determination of 0.91. The results demonstrated the critical roles of both particle ratio and data-driven models in wear performance analysis.

 Bu çalışmada, cam elyaf takviyeli epoksi kompozitlerin tribolojik performanslarını iyileştirmek amacıyla farklı miktarlarda 1-5 ağırlıkça eklenen Al2O3 partiküllerinin aşınma davranışları üzerindeki etkileri deneysel olarak araştırılmıştır. Elle yatırma yöntemiyle üretilen kompozit laminantlar, ball-on-disc bilye test düzeneği kullanılarak aşınma testlerine tabi tutulmuştur. Test edilen tüm numuneler arasında, 3 ağırlıkça Al2O3 içeren kompozit en yüksek aşınma direncini göstermiştir. Saf kompozitle karşılaştırıldığında, özgül aşınma oranı 70e kadar azalmıştır. Buna karşılık, 4 ve 5 Al2O3 ilavesi, partikül aglomerasyonu nedeniyle aşınma direncinde bir azalmaya neden olmuştur. En yüksek özgül aşınma oranı 260x10⁶ mm³Nm iken, 3 eklenen numunede bu değer 80x10⁶ mm³Nmye düşmüştür. Ayrıca, yapay sinir ağı ve farklı makine öğrenimi regresörleri kullanılarak aşınma oranı tahminleri gerçekleştirilmiştir. En düşük MAPE değerlerine Random Forest 17.62, Ridge regresörü 18.46 ve ANN 19.92 ulaşmış olup, bu da Al2O3 takviyeli cam elyaf kompozitler için güçlü tahmin performansına işaret etmektedir. Grid search metodu ile optimize edilen yapay sinir ağı modeli 0.9lık bir ortalama karesel hata ve 0.92lik bir belirlilik katsayısı değeri elde ederken, rastgele orman regresörü 0.91lik bir belirlilik katsayısı değeriyle güçlü bir genelleme göstermiştir. Sonuçlar, aşınma performansı analizinde hem parçacık oranının hem de veri odaklı modellerin kritik rollerini ortaya koymuştur.

Article Details

Experimental investigation of wear behavior in Al2O3-reinforced glass fiber composites and comparative analysis of artificial neural network and machine learning models

Authors:

Raşit Koray Ergün | İsmail Bayar | Hüseyin Köse

Publisher:

Nigde Omer Halisdemir University

Main Topics:

Makine Öğrenme (Diğer) | Triboloji | Kompozit ve Hibrit Malzemeler

Journal:

Niğde Ömer Halisdemir Üniversitesi Mühendislik Bilimleri Dergisi Vol.14 Issue 4, pp.1571-1581

Document Type:

Article

Publication Date:

10/2025

Document Language:

English

Document ID (Org No.):

2001360305

pISSN:

2564-6605

GFRP

Wear

Al2O3

Machine learning

ANNGFRP

Aşınma

Al2O3

Makine öğrenmesi

Yapay sinir ağları

Issues

Article Details

Authors: Raşit Koray Ergün | İsmail Bayar | Hüseyin Köse

Publisher: Nigde Omer Halisdemir University

Main Topics: Makine Öğrenme (Diğer) | Triboloji | Kompozit ve Hibrit Malzemeler

Journal: Niğde Ömer Halisdemir Üniversitesi Mühendislik Bilimleri Dergisi Vol.14 Issue 4, pp.1571-1581

Document Type: Article

Publication Date: 10/2025

Document Language: English

Document ID (Org No.): 2001360305

pISSN: 2564-6605

Issues

Authors:

Raşit Koray Ergün | İsmail Bayar | Hüseyin Köse

Publisher:

Nigde Omer Halisdemir University

Main Topics:

Makine Öğrenme (Diğer) | Triboloji | Kompozit ve Hibrit Malzemeler

Journal:

Niğde Ömer Halisdemir Üniversitesi Mühendislik Bilimleri Dergisi Vol.14 Issue 4, pp.1571-1581

Document Type:

Article

Publication Date:

10/2025

Document Language:

English

Document ID (Org No.):

2001360305

pISSN:

2564-6605