Investigation of The Abrasive Wear Behavior of GFRC And CFRC with Different Parameters Using Taguchi And Artificial Neural Networks Method

Mehmet Emin Demir

Makale detayları

Investigation of The Abrasive Wear Behavior of GFRC And CFRC with Different Parameters Using Taguchi And Artificial Neural Networks Method

Yazarlar:

Mehmet Emin Demir

Yayıncı:

Gazi University

Ana Başlıklar:

Makine Mühendisliğinde Optimizasyon Teknikleri | Triboloji

Dergi:

Politeknik Dergisi Vol.28 Issue 1, pp.215-228

Belge Tipi:

Makale

Yayın Tarihi:

1/2025

Belge Dili:

İngilizce

Document ID (Org No.):

2001339499

pISSN:

2147-9429

Anahtar kelimeler:

Özet:

 Fiber-reinforced composites are increasingly being utilized in various sectors, including aerospace, maritime, electronic components, and in elements exposed to wear such as bolts, nuts, cams, and gaskets. This study aims to determine the optimal processing parameters in abrasive wear tests conducted under varying wear conditions on glass and carbon fiber-reinforced composites. Employing a mixed-level L36 Taguchi orthogonal experimental design, tests were conducted on a pin on disk apparatus under different loads, sliding distances, and speeds. The results indicated that the most significant parameters affecting the coefficient of friction COF and mass loss were the type of fiber and the load. It was observed that an increase in load, sliding distance, and speed augmented the COF and mass loss. Predictions of the coefficient of friction and mass loss were made using a model developed in Artificial Neural Networks ANN, and these predictions were compared with experimental results. The R2 overall regression values for COF and mass loss in ANN were calculated as 0.98939 and 0.98349, respectively. ANN was found to provide more consistent results in predicting COF and mass loss compared to the Taguchi method.

 Fiber-reinforced composites are increasingly being utilized in various sectors, including aerospace, maritime, electronic components, and in elements exposed to wear such as bolts, nuts, cams, and gaskets. This study aims to determine the optimal processing parameters in abrasive wear tests conducted under varying wear conditions on glass and carbon fiber-reinforced composites. Employing a mixed-level L36 Taguchi orthogonal experimental design, tests were conducted on a pin on disk apparatus under different loads, sliding distances, and speeds. The results indicated that the most significant parameters affecting the coefficient of friction COF and mass loss were the type of fiber and the load. It was observed that an increase in load, sliding distance, and speed augmented the COF and mass loss. Predictions of the coefficient of friction and mass loss were made using a model developed in Artificial Neural Networks ANN, and these predictions were compared with experimental results. The R2 overall regression values for COF and mass loss in ANN were calculated as 0.98939 and 0.98349, respectively. ANN was found to provide more consistent results in predicting COF and mass loss compared to the Taguchi method.

 Fiber-reinforced composites are increasingly being utilized in various sectors, including aerospace, maritime, electronic components, and in elements exposed to wear such as bolts, nuts, cams, and gaskets. This study aims to determine the optimal processing parameters in abrasive wear tests conducted under varying wear conditions on glass and carbon fiber-reinforced composites. Employing a mixed-level L36 Taguchi orthogonal experimental design, tests were conducted on a pin on disk apparatus under different loads, sliding distances, and speeds. The results indicated that the most significant parameters affecting the coefficient of friction COF and mass loss were the type of fiber and the load. It was observed that an increase in load, sliding distance, and speed augmented the COF and mass loss. Predictions of the coefficient of friction and mass loss were made using a model developed in Artificial Neural Networks ANN, and these predictions were compared with experimental results. The R2 overall regression values for COF and mass loss in ANN were calculated as 0.98939 and 0.98349, respectively. ANN was found to provide more consistent results in predicting COF and mass loss compared to the Taguchi method.

 Fiber takviyeli kompozitler, havacılık, denizcilik, elektronik bileşenler ve cıvata, somun, kam ve conta gibi aşınmaya maruz kalan elemanlar gibi çeşitli sektörlerde giderek daha fazla kullanılmaktadır. Bu çalışma, cam ve karbon fiber takviyeli kompozitler üzerinde farklı aşınma koşullarında yapılan aşındırıcı aşınma testlerinde optimal işleme parametrelerini belirlemeyi amaçlamaktadır. Karışık düzeyli L36 Taguchi ortogonal deneysel bir tasarım kullanılarak, farklı yükler, kayma mesafeleri ve hızlar altında bir pim üzerinde disk cihazında testler gerçekleştirilmiştir. Sonuçlar, sürtünme katsayısı COF ve kütle kaybını etkileyen en önemli parametrelerin fiber tipi ve yük olduğunu göstermiştir. Yükün, kayma mesafesinin ve hızın artmasıyla COF ve kütle kaybının arttığı gözlemlenmiştir. Sürtünme katsayısı ve kütle kaybı için yapay sinir ağları ANN ile geliştirilen bir model kullanılarak tahminlerde bulunulmuş ve bu tahminler deneysel sonuçlarla karşılaştırılmıştır. ANNdeki COF ve kütle kaybı için R2 genel regresyon değerleri sırasıyla 0.98939 ve 0.98349 olarak hesaplanmıştır. ANNnin COF ve kütle kaybını tahmin etmede Taguchi yöntemine göre daha tutarlı sonuçlar sağladığı bulunmuştur.

Makale detayları

Investigation of The Abrasive Wear Behavior of GFRC And CFRC with Different Parameters Using Taguchi And Artificial Neural Networks Method

Yazarlar:

Mehmet Emin Demir

Yayıncı:

Gazi University

Ana Başlıklar:

Makine Mühendisliğinde Optimizasyon Teknikleri | Triboloji

Dergi:

Politeknik Dergisi Vol.28 Issue 1, pp.215-228

Belge Tipi:

Makale

Yayın Tarihi:

1/2025

Belge Dili:

İngilizce

Document ID (Org No.):

2001339499

pISSN:

2147-9429

GFRC

CFRC

Abrasive wear

Taguchi

ANNGFRC

CFRC

Abrasive aşınma

Taguchi

ANN

Sayı

Makale detayları

Yazarlar: Mehmet Emin Demir

Yayıncı: Gazi University

Ana Başlıklar: Makine Mühendisliğinde Optimizasyon Teknikleri | Triboloji

Dergi: Politeknik Dergisi Vol.28 Issue 1, pp.215-228

Belge Tipi: Makale

Yayın Tarihi: 1/2025

Belge Dili: İngilizce

Document ID (Org No.): 2001339499

pISSN: 2147-9429

Sayı

Yazarlar:

Mehmet Emin Demir

Yayıncı:

Gazi University

Ana Başlıklar:

Makine Mühendisliğinde Optimizasyon Teknikleri | Triboloji

Dergi:

Politeknik Dergisi Vol.28 Issue 1, pp.215-228

Belge Tipi:

Makale

Yayın Tarihi:

1/2025

Belge Dili:

İngilizce

Document ID (Org No.):

2001339499

pISSN:

2147-9429