Publication: Poli (Etilen-Oksit) (PEO) Kil Nanokompozit Malzemelerin Isısal ve Termomekanik Özelliklerinin Deneysel Olarak Saptanması ve Yapay Sinir Ağları Kullanılarak Simülasyonu
Loading...
Date
Authors
Journal Title
Journal ISSN
Volume Title
Abstract
Ağırlıkça %0, 5, 10, 15 ve 20 kil ihtiva eden Poli (etilen-oksit)/kil nanokompozitleri (PKN)solüsyon yardımıyla interkalasyona tabi tutuldu. Bu nanokompozitler ve saf Poli(etilen-oksit) (PEO); termogravimetrik analiz (TGA), dinamik mekanik analiz (DMA) ve diferansiyel taramalı kalorimetre (DSC) yöntemleri ile analiz edildi. Yapay sinir ağları (YSA) tekniği ileri beslemeli geri yayılım algoritması kullanılarak PKN'ne ısısal kararlılık, kristalleşme ve termomekanik özelliklerin saptanabilmesi amacıyla uygulandı.TGA, DMA ve DSC analizlerinin sonuçları ile hem üretilmemiş olan %1, 2, 3, 4, 6, 7, 8, 9, 11, 12, 13, 14, 16, 17, 18 ve 19 kil içerikli numuneler hem de üretilmiş nanokompozitlere ait sonuçlar YSA ile simüle edildi. Bu simülasyon sonuçları kullanılarak bozunma sıcaklığı, kül verimi, ergime ısı akışı, depolama modülü ve tan ? sonuçları hesaplandı.TGA, DSC ve DMA (hem depolama modülü (E?) hem de faz açısı (tan ?)) analizlerine ait simülasyon sonuçlarının deneysel verileri ile karşılaştırıldığı grafikleri çizildi. Simülasyon sonuçlarının deneysel sonuçlar ile iyi uyum içinde olduğu gözlemlendi. Simülasyon sonuçları dikkate alınarak ısısal kararlılık ve termomekanik davranışların arasındaki ilişkiler detaylıca incelendi.Deneysel veriler kullanılarak yapılan modelleme sonrasındaki simülasyon sonuçları gösterdi ki eğitimde kullanılan veri sayısı YSA'nın tahmin kabiliyetini etkileyen önemli bir faktördür. Ayrıca modelleme sonuçları ile YSA tekniğinin PKN'nin yapı-özellik ilişkilerini incelemek için potansiyel bir matematiksel araç olduğu sonucuna varılmıştır.
PEO/clay nanocomposites included 0, 5, 10, 15 and 20 wt.% clay were intercalated by solution method. The nanocomposites and pure PEO were characterized by dynamic mechanical analysis (DMA), differential scanning calorimetry (DSC) and thermogravimetric analysis (TGA) techniques. The artificial neural network (ANN) technique with a feed-forward back propagation algorithm was used to examine the thermal stability, crystallinity and thermomechanical properties of poly(ethylene-oxide)/clay nanocomposites.Both non-produced %1, 2, 3, 4, 6, 7, 8, 9, 11, 12, 13, 14, 16, 17, 18 and 19 clay included composites and the produced nanocomposites? DMA, DSC and TGA results were simulated by the experimental results. Decomposition temperature, char yield, enthalpy of melting, storage modulus and tan ? were successfully calculated by well-trained by the simulation results.Graphs of the predicted TGA, DSC and DMA (both storage modulus, E? and tan ? data) as a function of sample composition and experimental testing conditions were generated. The simulated data is in good agreement with experimental data. Based on simulation results, the relationship between thermal stability, crystallinity and thermomechanical properties was investigated in detail. Simulation results showed that the number of training dataset is one of the key factors in the prediction ability of ANN network. It was also confirmed that the ANN technique is a potential mathematical tool in the structure-property analysis of polymer/clay nanocomposites.
PEO/clay nanocomposites included 0, 5, 10, 15 and 20 wt.% clay were intercalated by solution method. The nanocomposites and pure PEO were characterized by dynamic mechanical analysis (DMA), differential scanning calorimetry (DSC) and thermogravimetric analysis (TGA) techniques. The artificial neural network (ANN) technique with a feed-forward back propagation algorithm was used to examine the thermal stability, crystallinity and thermomechanical properties of poly(ethylene-oxide)/clay nanocomposites.Both non-produced %1, 2, 3, 4, 6, 7, 8, 9, 11, 12, 13, 14, 16, 17, 18 and 19 clay included composites and the produced nanocomposites? DMA, DSC and TGA results were simulated by the experimental results. Decomposition temperature, char yield, enthalpy of melting, storage modulus and tan ? were successfully calculated by well-trained by the simulation results.Graphs of the predicted TGA, DSC and DMA (both storage modulus, E? and tan ? data) as a function of sample composition and experimental testing conditions were generated. The simulated data is in good agreement with experimental data. Based on simulation results, the relationship between thermal stability, crystallinity and thermomechanical properties was investigated in detail. Simulation results showed that the number of training dataset is one of the key factors in the prediction ability of ANN network. It was also confirmed that the ANN technique is a potential mathematical tool in the structure-property analysis of polymer/clay nanocomposites.
Description
Tez (yüksek lisans) -- Ondokuz Mayıs Üniversitesi, 2012
Libra Kayıt No: 71866
Libra Kayıt No: 71866
Citation
WoS Q
Scopus Q
Source
Volume
Issue
Start Page
End Page
84
